Subset (5)

F1000Journal · 8.11k rows

Split (1)

data · 8.11k rows

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/3-222/v1	15 Sep 14	{ "type": "Review", "title": "A review of creatine supplementation in age-related diseases: more than a supplement for athletes", "authors": [ "Rachel N. Smith", "Amruta S. Agharkar", "Eric B. Gonzales", "Rachel N. Smith", "Amruta S. Agharkar" ], "abstract": "Creatine is an endogenous compound synthesized from arginine, glycine and methionine. This dietary supplement can be acquired from food sources such as meat and fish, along with athlete supplement powders. Since the majority of creatine is stored in skeletal muscle, dietary creatine supplementation has traditionally been important for athletes and bodybuilders to increase the power, strength, and mass of the skeletal muscle. However, new uses for creatine have emerged suggesting that it may be important in preventing or delaying the onset of neurodegenerative diseases associated with aging. On average, 30% of muscle mass is lost by age 80, while muscular weakness remains a vital cause for loss of independence in the elderly population. In light of these new roles of creatine, the dietary supplement’s usage has been studied to determine its efficacy in treating congestive heart failure, gyrate atrophy, insulin insensitivity, cancer, and high cholesterol. In relation to the brain, creatine has been shown to have antioxidant properties, reduce mental fatigue, protect the brain from neurotoxicity, and improve facets/components of neurological disorders like depression and bipolar disorder. The combination of these benefits has made creatine a leading candidate in the fight against age-related diseases, such as Parkinson’s disease, Huntington’s disease, amyotrophic lateral sclerosis, long-term memory impairments associated with the progression of Alzheimer’s disease, and stroke. In this review, we explore the normal mechanisms by which creatine is produced and its necessary physiology, while paying special attention to the importance of creatine supplementation in improving diseases and disorders associated with brain aging and outlining the clinical trials involving creatine to treat these diseases.", "keywords": [ "Creatine", "central nervous system", "age-related diseases" ], "content": "Introduction\n\nThe usage of dietary supplementation in the United States is a multibillion-dollar industry, where creatine (N-[aminoiminomethyl]-N-methyl glycine) accounts for over 4 million kg and $200 million annually1,2. Creatine is an endogenous molecule found in all cells in the body and is synthesized in the kidney, liver, and pancreas using the amino acids arginine, glycine and methionine before entering the bloodstream3–5. From the plasma, creatine is transported into the cells via the creatine transporter protein (CRT)6,7. This transporter is critical for the distribution of creatine throughout the cells as well as for traversing the blood brain barrier (BBB), giving creatine access to the central nervous system (CNS).\n\nNearly 95% of creatine stores reside in skeletal muscle with the remaining 5% found in the brain, liver, testes, and kidneys8. Perhaps the most well understood role of creatine in physiology is its participation in energy production. More specifically, creatine maintains the intracellular levels of adenosine triphosphate (ATP) in skeletal muscle. This source of ATP is produced via oxidative phosphorylation, which is regulated by the mitochondria9. Within just a few seconds, muscle contraction utilizes the entire ATP store (2–5 mM) found in skeletal muscle10. ATP is regenerated using the phosphocreatine system where phosphocreatine donates its phosphate group to adenosine diphosphate (ADP) to form ATP. This reaction occurs rapidly and reversibly via the enzyme creatine kinase (CK), making the ATP replenishing capacity of both phosphocreatine and creatine kinase high. Conversely, at rest, ATP donates a phosphate group to creatine in order to replenish phosphocreatine stores for future muscle contraction use.\n\nAlthough primarily associated with energy production, mitochondria play an important role in the production of reactive oxygen species, dysregulation of calcium, excitotoxicity, and premature cellular death11–14. Likewise, creatine has important implications in antioxidant mechanisms, controlling intracellular calcium concentrations, regulating extracellular glutamate concentrations, and preventing the opening of the mitochondrial permeability transition pore (MPT)15–20. With evidence for creatine’s critical role in cellular bioenergetics, the phosphocreatine system in energy buffering, and the aforementioned implications in mechanisms associated with mitochondrial dysregulation, it is no surprising that creatine is the subject of investigation for improving the status of patients with neurodegenerative diseases that either result or progress by some mechanism of energy insufficiency.\n\n\nPhysiological creatine concentrations and creatine supplementation\n\nThere is a maximum capacity for the synthesis of endogenous creatine. To increase these levels, patients and athletes turn to creatine supplementation. These individuals that take in foods rich in creatine tend to have higher creatine levels21,22. The transport of creatine into cells is limited, since the capacity of creatine transport within each muscle cell is only 160 mmol/kg23. The possible beneficial effects of creatine are negligible if the creatine transporter is not functioning or if the maximal concentration of creatine within the cell has been reached. Evidence suggests that additives with creatine supplementation like proteins, carbohydrates, alpha lipoic acid, and D-pinitol can stimulate the movement of creatine into the cell, making creatine an ideal supplement for athletes with increased protein and carbohydrate intake24–27.\n\nIn general, a 70 kg human has a total creatine pool of 120 grams with 2 grams per day production from both dietary and endogenous sources8,28. Like many other supplements, supplementation reduces the normal physiological creatine production. This reduction is reversible as creatine supplementation is terminated8. Athletes use creatine supplementation to increase creatine phosphate stores. Elevated phosphocreatine leads to the phosphorylation of ADP to ATP and aids in limiting energy depletion during rapid muscle movement. Multiple studies have indicated significant improvements in sprint performance, body mass, fat-free body mass, weightlifting volumes, oxygen uptake and overall exercise performance following creatine supplementation29–40. Creatine loading in athletes can require 20 grams per day of supplementation, while maintenance dosing is roughly 5 grams per day. These studies dosed the subjects in a similar fashion. Serum creatine levels reached 2.17 mM and 0.8 mM at 2.5 hours following a 20 gram and 5 gram creatine bolus, respectively41,42. Creatine is excreted in the urine as creatinine with a daily turnover of 2 grams per day. Although creatine supplementation results in reduced, but reversible natural creatine production, creatine supplementation appears to have few unwanted side effects43. Thus, creatine is an attractive dietary supplement for athletes.\n\nPrior to the usage of creatine as an athletic enhancer, creatine has been the focus of research to understand the dietary supplement’s role in physiology for 150 years. Creatine supplementation became popular during the Barcelona Olympic Games as it was shown to enhance athletic performance44. Around the same time, two studies showed that creatine enhanced exercise performance via oral creatine ingestion23,45. With a clear understanding of the creatine/phosphocreatine system and its relation to the ADP/ATP energy metabolism in the mitochondria, studies began to shift their focus to understanding creatine’s role in pathophysiological conditions.\n\nIn addition to athletic performance, creatine usage has expanded to treat pathophysiological conditions including gyrate atrophy, post-stroke depression, congestive heart failure, chronic musculoskeletal pain disorders, atherosclerotic diseases and cisplatin nephrotoxicity28,46,47. Furthermore, a recent review proposed prophylactic creatine supplementation could reduce chances of preterm labor or hypoxic-ischemic encephalopathy48. Extensive research has demonstrated that the availability of phosphocreatine plays a role in skeletal muscle pathology and the associated pain can be alleviated by the intake of exogenous creatine49. New studies indicate that creatine plays a role in age-related neurological diseases and reduced brain functionality associated with Parkinson’s disease, Huntington’s disease, amyotrophic lateral sclerosis (ALS), long-term memory deficits, Alzheimer’s disease, and stroke. In the subsequent text, we will discuss creatine’s role in these neurodegenerative conditions.\n\n\nBenefits of creatine supplementation in age - associated declines in the brain\n\nAging is associated with lower levels of creatine and phosphocreatine, specifically in the skeletal muscle. Phosphocreatine regeneration rates following exercise fall approximately 8% each decade after age 3050. Creatine supplementation increases both creatine and phosphocreatine from 10–40% in athletes50. Furthermore, a recent review described a meta-analysis of the role creatine supplementation and resistance training plays on muscle health in an aging population. Based on the analysis of 13 published, creatine had an overall beneficial effect on aged individuals muscle mass51. Since the creatine transporter can readily transport creatine from the bloodstream across the BBB, it is reasonable to suggest that exogenous supplementation of creatine would increase concentrations in the brain, where endogenous creatine levels may be diminished as a person ages. In neurodegenerative disorders (as outlined below), creatine may help slow the progression of each condition.\n\nParkinson’s disease (PD) is a neurodegenerative disorder resulting from the loss of dopamine neurons in the midbrain with symptoms becoming apparent when approximately 60% of these neurons are lost52. Notable symptoms of PD include resting tremor, postural instability, bradykinesia, loss of muscle mass, strength, and increased ability to fatigue. The treatment measures for PD involve early detection of the disease and understanding how to slow PD progression once symptoms have been reported. Rodent models often used for the study of PD are induced by toxins, such as 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), to mimic the pathogenesis and progression of the disease marked by dopaminergic neuron loss, mitochondrial dysfunction, and oxidative stress53,54. In particular, complex I of the electron transport chain within the mitochondria is deficient in patients with Parkinson’s disease. The fact that postural instability, loss of muscle mass, and strength all occur in the progression of PD, and are coupled with creatine’s ability to alter cellular energetics, has led to the hypothesis that the creatine dietary supplementation could minimize the associated symptoms with Parkinson’s disease.\n\nAn early study testing the benefit of creatine in MPTP-induced Parkinson’s disease mice showed significant neuroprotection with 1% creatine supplementation in diet55. In 2006, a group of investigators at the National Institute of Neurological Disorders and Stroke (NINDS) began a phase III clinical trial for creatine in 200 patients affected by Parkinson’s disease after second phase preliminary data showed that creatine was able to slow down the progression of the disease56. A year later, a double-blind study compared the control group (no creatine supplementation) to the test group (20 Parkinson’s disease patients), which received a creatine loading dose of 20 grams per day for 5 days and a creatine maintenance dose of 5 grams per day thereafter57. The purpose of this study was to specifically explore if creatine could help increase muscle strength in idiopathic Parkinson’s disease patients. Both groups received resistance training during the study. A difference was observed in the creatine supplemented group with some of the strength exercises used versus the control group57. In September 2013, the NINDS announced that the phase III clinical trial for creatine use in Parkinson’s disease was halted because the study would result in an observable significant difference58. The test subjects tolerance for creatine or creatine associated side effects were not the cause for stopping the trial. The patients in this creatine clinical trial received 10 grams creatine daily for up to 5 years. Although this outcome is disappointing, the ability for creatine to alter energy dysfunction in addition to muscle strength may still have a combinatory effect with other Parkinson’s disease drug treatments. The possibility remains that creatine may be beneficial for Parkinson’s disease patients, but more work needs to be done to demonstrate the dietary supplements efficacy.\n\nHuntington’s disease (HD) is a neurodegenerative disease where the onset of symptoms occurs in midlife. Once the symptoms begin, a patient can expect to live, on average, 20 more years59. Those that develop Huntington’s disease possess genetically inherited mutations in the number of cytosine-adenine-guanine (CAG) repeats in the huntingtin gene responsible for producing the huntingtin protein60–63. The huntingtin protein is expressed throughout the central and peripheral nervous systems. Upon the onset of HD symptoms, the patient begins to exhibit changes in mood, cognition, and motor coordination60–63. In addition to contributing to an abnormal gait, resting tremor and even epileptic seizures associated with Huntington’s disease, the mutant huntingtin protein product leads to impaired energy metabolism64. Without a cure for Huntington’s disease, the impairment of energy metabolism offers an avenue and target for new therapies. Furthermore, there is an observed reduction in the phosphocreatine and inorganic phosphate ratio in Huntington’s disease patients’ muscle tissue, which may indicate the Huntington mutation’s involvement in dysregulating the phosphocreatine/creatine ratio65. In a mitochondrial toxin Huntington’s disease mutant mouse model, R6/2, creatine is hypothesized to act as a means of buffering, or providing a larger phosphocreatine pool for rapid conversion of ADP to ATP, energy within the cell5,66–68. Furthermore, as an extension of the work done by Matthews and colleagues68, the Ferrante group69 showed that lifespan was extended by 9.4, 17.7 and 4.4% when supplemented with 1, 2, or 3% creatine in the diet, respectively69.\n\nSeveral Huntington’s disease transgenic mice strains have been developed to study the dysfunction in energy metabolism, the electron transport enzymes in the mitochondria, and excessive excitotoxicity associated with the disease53,70–74. Supplementation with exogenous creatine in one transgenic mouse model showed improved motor performance, reduced atrophy of neurons, and huntingtin protein aggregates, and an observed increased survival rate or life-span17,69,75. Following these studies, a phase II clinical trial ensued to assess creatine tolerability given at a dosage of 8 grams per day to Huntington’s disease patients76. From these studies, Hersch and colleagues determined that 8-hydroxy-2’-deoxyguanosine (8OH2’dG), a marker for damaged DNA, was abnormally high in patients with HD, but was reduced after creatine treatment76. In an interview regarding the phase III clinical trial of creatine usage in Huntington’s disease patients, the CREST-E (Creatine Safety, Tolerability, and Efficacy) clinical trial, Hersch reported that 8OH2’dG had returned to normal levels. Subsequently, there was an observed reduction in brain deterioration rate when patients were supplemented with creatine and that creatine kinase is a potential biomarker for HD77. The potential of creatine as a viable therapy for HD remains to be seen and the results from CREST-E clinical trial will provide some indication to the dietary supplement’s utility. Thus, creatine supplementation remains a potential therapy for Huntington’s disease, however further studies are needed.\n\nAmyotrophic lateral sclerosis (ALS), or Lou Gehrig’s disease, is marked by the loss of voluntary muscle control from the progressive degeneration of motor neurons78,79 with subsequent neuronal loss resulting in paralysis80,81. The cause and cure for ALS remain elusive. The most promising treatment is the drug riluzole, which only increases the lifespan of those with the disease by 6 months82. To complicate the search for effective treatments, as many as 50% of ALS patients experience cognitive impairment that is revealed when they undergo specialized testing for neuropsychological deficits83. In addition to motor neuron loss observed in ALS patients, cognitive impairment most often associated with the frontotemporal region of the brain is not always present83. Furthermore, despite identifying a genetic overlap in ALS and other neurodegenerative disease mechanisms, the sporadic occurrence of cognitive impairment in ALS patients obscures understanding a clear etiology of the disease83.\n\nAt a molecular level, ALS is characterized by altered glutamate homeostasis, oxidative damage, elevated intracellular calcium concentrations, mitochondrial swelling, and electron transport chain complex I deficiencies leading to reduced energy intake84–87. Generally, mutations in the gene responsible for the production of the enzyme, superoxide dismutase (SOD1) are common in many cases of familial ALS80. The loss of function associated with SOD1 mutations reported in ALS translates to the accumulation of toxic free radicals from superoxide generated by the mitochondria80,88,89. This suggests that the build up of free radicals results in altered energy production. Thus, creatine may serve as an energy alternative that is beneficial for ALS patients.\n\nKlivenyi and colleagues studied transgenic mice with a mutated human SOD1 gene and assessed the neuroprotective effects of creatine. This was in response to the promotion of survival and improved motor coordination they observed with long-term creatine supplementation90. Along with the proposed creatine benefits in protecting neurons from insufficient energy production, the results indicated that creatine administration protected neurons from oxidative damage. In contrast, two completed clinical trials in 2003 and 2004 tested oral creatine supplementation and provided little notable improvements in lifespan, muscle strength, or motor unit numbers in patients with ALS91,92. Although there is an observed trend toward enhanced survival following creatine supplementation, these studies distinguished between large differences (30–50% difference). Due to the high threshold for observing significant differences, there remains a possiblilty that creatine has a subtle effect. Currently, the long-term effects of creatine supplementation are being studied in a phase II clinical trial associated with the Northeast Amyotrophic Lateral Sclerosis Consortium (NEALS).\n\nThe ability to encode new memories (working memory), recall previous events (episodic/long-term/declarative memory) and short-term (primary/active) memory declines with age93,94. Previous reports indicated that phosphocreatine stores are quickly depleted upon brain activation, while ATP concentrations remain constant95,96. A reduction in creatine levels may have an effect on the immediate recall of knowledge based on the dramatic drop in phosphocreatine levels from brain activation. In 2003, Rae and colleagues questioned if creatine supplementation enhanced intelligence in healthy subjects97. For 6 weeks, subjects were given 5 grams of creatine orally per day. Following this protocol, the subjects showed improvements in working memory and intelligence utilizing the backward digit span and Raven’s Advanced Progressive Matrices tasks, respectively. Furthermore, each of the participants were vegetarians, which supported the role that exogenous creatine has on increasing participants’ serum creatine levels. Using magnetic resonance spectroscopy to measure creatine levels, it was determined that creatine levels increased in the brain, solidifying that creatine is not solely distributed in skeletal muscle98. Creatine supplementation has been considered for the improvement of memory in the elderly. One such study in the United Kingdom reported that subjects with an average age of 76 saw improvements in long-term memory when supplemented with 20 grams per day of creatine for 1 week99. In addition to improvement in the long-term memory task, the elderly subjects improved in both forward and backward spatial recall as well as forward number recall. Based on these results, the investigators concluded that creatine could enhance cognition in elderly subjects, although follow up studies have not elucidated a mechanism by which creatine does this. Currently, there are few studies focused on the role that creatine supplementation plays on cognition and memory. In addition, how creatine improves memory in the aforementioned studies is yet to be understood at the molecular and cellular levels.\n\nAlzheimer’s disease (AD) is a neurodegenerative disease marked by neurofibrillary plaques and tangles in the brain. One of the challenging characteristics of AD is the inability to definitively determine if a patient has the disease while alive. Only during the postmortem exam the disease can be definitively diagnosed100. However, these hallmarks can be observed in the postmortem brains of individuals that did not display dementia or deteriorated cognitive function. As the disease progresses, the symptoms include severe dementia, confusion, and the loss of long-term memory. Although the onset of AD is not entirely understood, studies have shown that high-energy metabolism precedes the onset of AD while there are increased levels of myo-inositol, an important structural component of lipids, in comparison to the relative creatine concentrations. Furthermore, this increase in the ratio of myo-inositol and creatine precedes the onset of dementia in individuals with Down’s syndrome101–103. Creatine was shown to be protective of rat hippocampal neurons when confronted with beta-amyloid (Aβ) toxicity104. One possible mechanism to intervene with the progression of Alzheimer’s disease is creatine kinase. Creatine kinase is responsible for the conversion of ATP to ADP and vice versa105 and tends to be susceptible to high levels of oxidative damage in the brains of Alzheimer’s disease patients105. In Alzheimer’s disease, creatine kinase activity is reduced by as much as 86% along with a reduction of in creatine kinase protein expression of 14%, which suggests that the Alzheimer’s disease brain has lower levels of phosphocreatine in the beginning stages of the disease101,106. Although creatine kinase levels were reduced, studies have questioned the involvement of creatine upon finding deposits of the molecule in amyloid precursor protein (APP) in transgenic mice107. The solubility of creatine in an aqueous solution is 100 mM, however the total creatine concentration in the brain only reaches 20 mM10. Possible explanations for the origin of these creatine deposits in the transgenic mice models include: (1) spillage from neuronal cell death, (2) excess oligodendrocyte production of creatine, (3) limited creatine uptake by the CRT and (4) the oxidation of creatine kinase that limits the formation of phosphocreatine108–111. Although each of these possibilities has been studied extensively, it remains unclear to the exact origin of the creatine deposits. This further reiterates that the exact role of creatine in AD is still yet to be understood and may be more complicated than previously thought.\n\nStroke is “defined as an acute neurologic dysfunction of vascular origin with sudden (within seconds) or a least rapid (within hours) occurrence of symptoms and signs corresponding to the involvement of focal areas in the brain”112. In the past decade, stroke has been the second leading cause of death113. Ischemic stroke, which is more common114,115, occurs when the brain is deprived of glucose and oxygen due to insufficient blood supply. This causes acidosis, an increase in intracellular calcium, and the formation of reactive oxygen species leading to ischemic cell death116. Symptoms of ischemic stroke include sudden numbness on one side of the body, vision disturbances, difficulty in speaking and understanding, imbalance, and loss of coordination. The available treatments for ischemic stroke consist of the use of tissue plasminogen activator (tPA)117 and/or surgical treatments. The commonly used in vitro model for ischemic stroke involves oxygen-glucose deprivation118,119 and rodent model for stroke involves mechanical or thromboembolic occlusion of cerebral vessel to mimic cerebral ischemia120–123.\n\nAn early study involving eight stroke patients has demonstrated that creatine and phosphocreatine content is reduced in the ischemic brain124. In rat hippocampal slices, pre-incubation with creatine (0.03–3 mmol/L) dose-dependently reduced damage due to anoxia125. Another in-vitro study in rat hippocampal slices showed dose dependent increase in phosphocreatine concentration and delay in anoxic depolarization after incubation with 1 mM creatine126. Creatine was shown to exert protective antioxidant effect in U937 human promonocytic cells after oxidative damage127. Creatine has been shown to be neuroprotective in an experimental model of anoxia in neonatal mice supplemented with creatine128–130. Also, rats subjected to creatine pretreatment before cerebral hypoxia showed a reduction (25%) in the volume of edematous brain tissue compared when compared to control131. Zhu et al. showed that mice supplemented with 2% creatine in the diet showed reduced neuronal damage compared to control groups following middle cerebral artery occlusion that causes ischemic stroke. The study indicated that this beneficial effect of creatine is due to the restoration of energy depletion and inhibition of caspase activation along with some other unknown mechanisms132. These studies support the fact that creatine could be a potential compound to be used as a prophylactic, or preventative, dietary supplement in patients at high risk for stroke133. Still, more work is needed to demonstrate the efficacy of creatine to prevent stroke and to develop a creatine supplementation regimen to help patients at risk for stroke to avoid the debilitating event.\n\n\nConclusions\n\nCreatine has the potential to elicit positive effects in muscle strength, memory, and has further influence on neurodegenerative conditions. It remains to be seen if creatine has the ability to alter age-associated, progressive neurodegenerative disorders once individuals are in intermediate or late stages of the disease. However, based on the creatine interaction with energy metabolism and subsequent neuroprotective mechanisms, the interest for studying alternative uses for creatine in physiology is enhanced. Unfortunately, the phase III clinical trial for Parkinson’s disease and creatine was halted. However, there are ongoing clinical trials for creatine. Creatine supplementation is in a phase III clinical trial for the treatment of the energetic deficiencies in Huntington’s disease. With promising effects thus far, CREST-E remains continually funded by the National Center for Complementary and Alternative Medicine (NCCAM) and the Food and Drug Administration (FDA) as the largest therapeutic trial ever for Huntington’s disease. The usefulness of this compound may prove an important deterrent at beginning stages of dementia or even for increasing muscle strength in Parkinson’s disease patients. Furthermore, creatine may serve as a preventative treatment for the long-term consequences of stroke but may play a more complicated role in Alzheimer’s disease. Despite the wide range of uses for creatine supplementation, this dietary supplement should be the focus of additional studies for the treatment of age-related diseases. Going forward, one must consider if there are other mechanisms for which creatine acts to be protective and beneficial. These alternative mechanisms and the molecular/cellular targets for creatine remain to be determined and fully characterized.", "appendix": "Author contributions\n\n\n\nRNS wrote the first draft of the manuscript, performed the initial literature search, and revised the manuscript. ASA contributed to the literature search, wrote sections of the manuscript, revised the manuscript, and formatted the references. EBG provided overall supervision, along with critical review and revision of the final manuscript. All authors have read and agreed to the publication of the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nRachel N. Smith is supported by a fellowship on an NIH National Institute on Aging Training Grant (T32AG020494). Eric B. Gonzales is supported by an American Heart Association Beginning Grant-in-Aid (12BGIA8820001) and an UNTHSC Institute for Aging and Alzheimer’s Disease Research (IAADR) Pilot Seed Grant.\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 members of the Gonzales laboratory for their helpful discussions.\n\n\nReferences\n\nZeisel SH: Regulation of “nutraceuticals”. Science. 1999; 285(5435): 1853–1855. PubMed Abstract \| Publisher Full Text\n\nSchnirring L: Creatine supplements face scrutiny: Will users pay later? Phys Sports Med. 1998; 26: 15–23.\n\nBloch K, Schoenheimer R: The Biological Precursors Of Creatine. J Biol Chem. 1941; 138(1): 167–194. Reference Source\n\nTarnopolsky MA, Beal MF: Potential for creatine and other therapies targeting cellular energy dysfunction in neurological disorders. Ann Neurol. 2001; 49(5): 561–574. PubMed Abstract \| Publisher Full Text\n\nAdhihetty PJ, Beal MF: Creatine and its potential therapeutic value for targeting cellular energy impairment in neurodegenerative diseases. Neuromolecular Med. 2008; 10(4): 275–290. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChristie DL: Functional insights into the creatine transporter. In: Gajja S. Salomons, M Wyss, ed. Creatine and creatine kinase in health and disease. Netherlands: Springer; 2007; 46: 99–118. Publisher Full Text\n\nLowe MT, Faull RL, Christie DL, et al.: The distribution of the creatine transporter throughout the human brain reveals a spectrum of creatine transporter immunoreactivity. J Comp Neurol. 2014. PubMed Abstract \| Publisher Full Text\n\nWalker JB: Creatine: biosynthesis, regulation, and function. Adv Enzymol Relat Areas Mol Biol. 1979; 50: 177–242. PubMed Abstract \| Publisher Full Text\n\nAdhihetty PJ, Irrcher I, Joseph AM, et al.: Plasticity of skeletal muscle mitochondria in response to contractile activity. Exp Physiol. 2003; 88(1): 99–107. PubMed Abstract \| Publisher Full Text\n\nWallimann T, Wyss M, Brdiczka D, et al.: Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: the ‘phosphocreatine circuit’ for cellular energy homeostasis. Biochem J. 1992; 281(Pt 1): 21–40. PubMed Abstract \| Free Full Text\n\nPeng TI, Greenamyre JT: Privileged access to mitochondria of calcium influx through N-methyl-D-aspartate receptors. Mol Pharmacol. 1998; 53(6): 974–980. PubMed Abstract\n\nGreen DR, Reed JC: Mitochondria and apoptosis. Science. 1998; 281(5381): 1309–1312. PubMed Abstract \| Publisher Full Text\n\nSteeghs K, Benders A, Oerlemans F, et al.: Altered Ca2+ responses in muscles with combined mitochondrial and cytosolic creatine kinase deficiencies. Cell. 1997; 89(1): 93–103. PubMed Abstract \| Publisher Full Text\n\nWyss M, Braissant O, Pischel I, et al.: Creatine and creatine kinase in health and disease--a bright future ahead? Subcell Biochem. 2007; 46: 309–334. PubMed Abstract \| Publisher Full Text\n\nXu CJ, Klunk WE, Kanfer JN, et al.: Phosphocreatine-dependent glutamate uptake by synaptic vesicles. A comparison with atp-dependent glutamate uptake. J Biol Chem. 1996; 271(23): 13435–13440. PubMed Abstract \| Publisher Full Text\n\nLawler JM, Barnes WS, Wu G, et al.: Direct antioxidant properties of creatine. Biochem Biophys Res Commun. 2002; 290(1): 47–52. PubMed Abstract \| Publisher Full Text\n\nDedeoglu A, Kubilus JK, Yang L, et al.: Creatine therapy provides neuroprotection after onset of clinical symptoms in Huntington’s disease transgenic mice. J Neurochem. 2003; 85(6): 1359–1367. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAraujo MB, Moura LP, Junior RC, et al.: Creatine supplementation and oxidative stress in rat liver. J Int Soc Sports Nutr. 2013; 10(1): 54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Andrade RB, Gemelli T, Rojas DB, et al.: Creatine and Pyruvate Prevent the Alterations Caused by Tyrosine on Parameters of Oxidative Stress and Enzyme Activities of Phosphoryltransfer Network in Cerebral Cortex of Wistar Rats. Mol Neurobiol. 2014. PubMed Abstract \| Publisher Full Text\n\nStefani GP, Nunes RB, Dornelles AZ, et al.: Effects of creatine supplementation associated with resistance training on oxidative stress in different tissues of rats. J Int Soc Sports Nutr. 2014; 11(1): 11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChevreul ME: Sur une nouvelle substance contenue dans la chair de boeuf. Paris Mus Hist Nat N Ann. 1832; I: 306–316.\n\nWyss M, Kaddurah-Daouk R: Creatine and creatinine metabolism. Physiol Rev. 2000; 80(3): 1107–1213. PubMed Abstract\n\nHarris RC, Soderlund K, Hultman E: Elevation of creatine in resting and exercised muscle of normal subjects by creatine supplementation. Clin Sci (Lond). 1992; 83(3): 367–374. PubMed Abstract\n\nGreen AL, Simpson EJ, Littlewood JJ, et al.: Carbohydrate ingestion augments creatine retention during creatine feeding in humans. Acta Physiol Scand. 1996; 158(2): 195–202. PubMed Abstract \| Publisher Full Text\n\nSteenge GR, Simpson EJ, Greenhaff PL: Protein- and carbohydrate-induced augmentation of whole body creatine retention in humans. J Appl Physiol (1985). 2000; 89(3): 1165–1171. PubMed Abstract\n\nBurke DG, Chilibeck Pd, Parise GA, et al.: The effect of alpha lipoic acid supplementation on resting muscle creatine during acute creatine loading. FASEB J. 2001; 15(5): A814.\n\nGreenwood M, Kreider RB, Almada AL, et al.: D-pinitol augments whole body creatine retention in man. J Exercise Physiol Online. 2001; 4(4): 41–47. Reference Source\n\nPersky AM, Brazeau GA: Clinical pharmacology of the dietary supplement creatine monohydrate. Pharmacol Rev. 2001; 53(2): 161–176. PubMed Abstract\n\nIzquierdo M, Ibanez J, Gonzalez-Badillo JJ, et al.: Effects of creatine supplementation on muscle power, endurance, and sprint performance. Med Sci Sports Exerc. 2002; 34(2): 332–343. PubMed Abstract\n\nMujika I, Padilla S, Ibanez J, et al.: Creatine supplementation and sprint performance in soccer players. Med Sci Sports Exerc. 2000; 32(2): 518–525. PubMed Abstract \| Publisher Full Text\n\nBecque MD, Lochmann JD, Melrose DR: Effects of oral creatine supplementation on muscular strength and body composition. Med Sci Sports Exerc. 2000; 32(3): 654–658. PubMed Abstract \| Publisher Full Text\n\nKreider RB, Ferreira M, Wilson M, et al.: Effects of creatine supplementation on body composition, strength, and sprint performance. Med Sci Sports Exerc. 1998; 30(1): 73–82. PubMed Abstract \| Publisher Full Text\n\nVandenberghe K, Goris M, Van Hecke P, et al.: Long-term creatine intake is beneficial to muscle performance during resistance training. J Appl Physiol (1985). 1997; 83(6): 2055–2063. PubMed Abstract\n\nVolek JS, Boetes M, Bush JA, et al.: Response of testosterone and cortisol concentrations to high-intensity resistance exercise following creatine supplementation. J Strength Conditioning Res. 1997; 11(3). Reference Source\n\nJacobs I, Bleue S, Goodman J: Creatine ingestion increases anaerobic capacity and maximum accumulated oxygen deficit. Can J Appl Physiol. 1997; 22(3): 231–243. PubMed Abstract \| Publisher Full Text\n\nDawson B, Cutler M, Moody A, et al.: Effects of oral creatine loading on single and repeated maximal short sprints. Aust J Sci Med Sport. 1995; 27(3): 56–61. PubMed Abstract\n\nMeir R: Practical application of oral creatine supplementation in professional rugby league: A case study. Australian Strength and Conditioning Coach. 1995; 3(3): 6–10. Reference Source\n\nAntonio J, Ciccone V: The effects of pre versus post workout supplementation of creatine monohydrate on body composition and strength. J Int Soc Sports Nutr. 2013; 10: 36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Salles Painelli V, Alves VT, Ugrinowitsch C, et al.: Creatine supplementation prevents acute strength loss induced by concurrent exercise. Eur J Appl Physiol. 2014; 114(8): 1749–1755. PubMed Abstract \| Publisher Full Text\n\nCamic CL, Housh TJ, Zuniga JM, et al.: The effects of polyethylene glycosylated creatine supplementation on anaerobic performance measures and body composition. J Strength Cond Res. 2014; 28(3): 825–833. PubMed Abstract \| Publisher Full Text\n\nMesa JL, Ruiz JR, Gonzalez-Gross MM, et al.: Oral creatine supplementation and skeletal muscle metabolism in physical exercise. Sports Med. 2002; 32(14): 903–944. PubMed Abstract \| Publisher Full Text\n\nPersky AM, Muller M, Derendorf H, et al.: Single- and multiple-dose pharmacokinetics of oral creatine. J Clin Pharmacol. 2003; 43(1): 29–37. PubMed Abstract \| Publisher Full Text\n\nHall M, Trojian TH: Creatine supplementation. Curr Sports Med Rep. 2013; 12(4): 240–244. PubMed Abstract \| Publisher Full Text\n\nAnderson O: Creatine propels british athletes to olympic gold medals: Is creatine the one true ergogenic aid? Running Research News. 1993; 9: 1–5.\n\nGreenhaff PL, Casey A, Short AH, et al.: Influence of oral creatine supplementation of muscle torque during repeated bouts of maximal voluntary exercise in man. Clin Sci (Lond). 1993; 84(5): 565–571. PubMed Abstract\n\nWyss M, Schulze A: Health implications of creatine: can oral creatine supplementation protect against neurological and atherosclerotic disease? Neuroscience. 2002; 112(2): 243–260. PubMed Abstract \| Publisher Full Text\n\nGenc G, Okuyucu A, Meydan BC, et al.: Effect of free creatine therapy on cisplatin-induced renal damage. Ren Fail. 2014; 36(7): 1108–1113. PubMed Abstract \| Publisher Full Text\n\nDickinson H, Ellery S, Ireland Z, et al.: Creatine supplementation during pregnancy: summary of experimental studies suggesting a treatment to improve fetal and neonatal morbidity and reduce mortality in high-risk human pregnancy. BMC Pregnancy Childbirth. 2014; 14: 150. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeader A, Amital D, Rubinow A, et al.: An open-label study adding creatine monohydrate to ongoing medical regimens in patients with the fibromyalgia syndrome. Ann N Y Acad Sci. 2009; 1173: 829–836. PubMed Abstract \| Publisher Full Text\n\nKreider RB: Effects of creatine supplementation on performance and training adaptations. Mol Cell Biochem. 2003; 244(1–2): 89–94. PubMed Abstract \| Publisher Full Text .\n\nCandow DG, Chilibeck PD, Forbes SC: Creatine supplementation and aging musculoskeletal health. Endocrine. 2014; 45(3): 354–361. PubMed Abstract \| Publisher Full Text\n\nLang AE, Lozano AM: Parkinson’s disease. First of two parts. N Engl J Med. 1998; 339(15): 1044–1053. PubMed Abstract \| Publisher Full Text\n\nBeal MF: Mitochondria, free radicals, and neurodegeneration. Curr Opin Neurobiol. 1996; 6(5): 661–666. PubMed Abstract \| Publisher Full Text\n\nThomas B, Beal MF: Parkinson’s disease. Hum Mol Genet. 2007; 16(R2): R183–94. Publisher Full Text\n\nMatthews RT, Ferrante RJ, Klivenyi P, et al.: Creatine and cyclocreatine attenuate MPTP neurotoxicity. Exp Neurol. 1999; 157(1): 142–149. PubMed Abstract \| Publisher Full Text\n\nNINDS NET-PD Investigators. A randomized, double-blind, futility clinical trial of creatine and minocycline in early Parkinson disease. Neurology. 2006; 66(5): 664–671. PubMed Abstract \| Publisher Full Text\n\nHass CJ, Collins MA, Juncos JL: Resistance training with creatine monohydrate improves upper-body strength in patients with Parkinson disease: a randomized trial. Neurorehabil Neural Repair. 2007; 21(2): 107–115. PubMed Abstract \| Publisher Full Text\n\nStatement on the termination of NET-PD LS-1 study. NET-PD NIH Exploratory Trials in Parkinson’s Disease Web site. 2013. Reference Source\n\nWalker FO: Huntington’s disease. Lancet. 2007; 369(9557): 218–228. PubMed Abstract \| Publisher Full Text\n\nStrong TV, Tagle DA, Valdes JM, et al.: Widespread expression of the human and rat Huntington’s disease gene in brain and nonneural tissues. Nat Genet. 1993; 5(3): 259–265. PubMed Abstract \| Publisher Full Text\n\nLandwehrmeyer GB, McNeil SM, Dure LS 4th, et al.: Huntington’s disease gene: regional and cellular expression in brain of normal and affected individuals. Ann Neurol. 1995; 37(2): 218–230. PubMed Abstract \| Publisher Full Text\n\nSharp AH, Loev SJ, Schilling G, et al.: Widespread expression of Huntington’s disease gene (IT15) protein product. Neuron. 1995; 14(5): 1065–1074. PubMed Abstract \| Publisher Full Text\n\nFerrante RJ, Gutekunst CA, Persichetti F, et al.: Heterogeneous topographic and cellular distribution of huntingtin expression in the normal human neostriatum. J Neurosci. 1997; 17(9): 3052–3063. PubMed Abstract\n\nRyu H, Ferrante RJ: Emerging chemotherapeutic strategies for Huntington’s disease. Expert Opin Emerg Drugs. 2005; 10(2): 345–363. PubMed Abstract \| Publisher Full Text\n\nKoroshetz WJ, Jenkins BG, Rosen BR, et al.: Energy metabolism defects in Huntington’s disease and effects of coenzyme Q10. Ann Neurol. 1997; 41(2): 160–165. PubMed Abstract \| Publisher Full Text\n\nWallimann T, Hemmer W: Creatine kinase in non-muscle tissues and cells. Mol Cell Biochem. 1994; 133–134: 193–220. PubMed Abstract \| Publisher Full Text\n\nO’Gorman E, Beutner G, Dolder M, et al.: The role of creatine kinase in inhibition of mitochondrial permeability transition. FEBS Lett. 1997; 414(2): 253–257. PubMed Abstract \| Publisher Full Text\n\nMatthews RT, Yang L, Jenkins BG, et al.: Neuroprotective effects of creatine and cyclocreatine in animal models of Huntington’s disease. J Neurosci. 1998; 18(1): 156–163. PubMed Abstract\n\nFerrante RJ, Andreassen OA, Jenkins BG, et al.: Neuroprotective effects of creatine in a transgenic mouse model of Huntington’s disease. J Neurosci. 2000; 20(12): 4389–4397. PubMed Abstract\n\nBrouillet E, Jenkins BG, Hyman BT, et al.: Age-dependent vulnerability of the striatum to the mitochondrial toxin 3-nitropropionic acid. J Neurochem. 1993; 60(1): 356–359. PubMed Abstract \| Publisher Full Text\n\nBeal MF, Brouillet E, Jenkins BG, et al.: Neurochemical and histologic characterization of striatal excitotoxic lesions produced by the mitochondrial toxin 3-nitropropionic acid. J Neurosci. 1993; 13(10): 4181–4192. PubMed Abstract\n\nSchulz JB, Beal MF: Neuroprotective effects of free radical scavengers and energy repletion in animal models of neurodegenerative disease. Ann N Y Acad Sci. 1995; 765: 100–10; discussion 116–8. PubMed Abstract \| Publisher Full Text\n\nBrowne SE, Ferrante RJ, Beal MF: Oxidative stress in Huntington’s disease. Brain Pathol. 1999; 9(1): 147–163. PubMed Abstract \| Publisher Full Text\n\nGrunewald T, Beal MF: Bioenergetics in Huntington’s disease. Ann N Y Acad Sci. 1999; 893: 203–213. PubMed Abstract \| Publisher Full Text\n\nAndreassen OA, Dedeoglu A, Ferrante RJ, et al.: Creatine increase survival and delays motor symptoms in a transgenic animal model of Huntington’s disease. Neurobiol Dis. 2001; 8(3): 479–491. PubMed Abstract \| Publisher Full Text\n\nHersch SM, Gevorkian S, Marder K, et al.: Creatine in huntington disease is safe, tolerable, bioavailable in brain and reduces serum 8OH2’dG. Neurology. 2006; 66(2): 250–252. PubMed Abstract \| Publisher Full Text\n\nKim J, Amante DJ, Moody JP, et al.: Reduced creatine kinase as a central and peripheral biomarker in Huntington’s disease. Biochim Biophys Acta. 2010; 1802(7–8): 673–681. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalton JN: Brain’s diseases of the nervous system. 10th ed. New York: Oxford University Press, 1993: 443–449. Reference Source\n\nDementia and motoneurone disease. Lancet. 1990; 335(8700): 1250–1252. Publisher Full Text\n\nHervias I, Beal MF, Manfredi G: Mitochondrial dysfunction and amyotrophic lateral sclerosis. Muscle Nerve. 2006; 33(5): 598–608. PubMed Abstract \| Publisher Full Text\n\nDugdale DC, Hoch DB, Zieve D: Amyotrophic lateral sclerosis. A D A M Medical Encyclopedia. 2010.\n\nZoccolella S, Beghi E, Palagano G, et al.: Riluzole and amyotrophic lateral sclerosis survival: a population-based study in southern italy. Eur J Neurol. 2007; 14(3): 262–268. PubMed Abstract \| Publisher Full Text\n\nPhukan J, Pender NP, Hardiman O: Cognitive impairment in amyotrophic lateral sclerosis. Lancet Neurol. 2007; 6(11): 994–1003. PubMed Abstract \| Publisher Full Text\n\nStrong M, Rosenfeld J: Amyotrophic lateral sclerosis: a review of current concepts. Amyotroph Lateral Scler Other Motor Neuron Disord. 2003; 4(3): 136–143. PubMed Abstract \| Publisher Full Text\n\nSiklos L, Engelhardt J, Harati Y, et al.: Ultrastructural evidence for altered calcium in motor nerve terminals in amyotropic lateral sclerosis. Ann Neurol. 1996; 39(2): 203–216. PubMed Abstract \| Publisher Full Text\n\nCarri MT, Ferri A, Battistoni A, et al.: Expression of a Cu,Zn superoxide dismutase typical of familial amyotrophic lateral sclerosis induces mitochondrial alteration and increase of cytosolic Ca2+ concentration in transfected neuroblastoma SH-SY5Y cells. FEBS Lett. 1997; 414(2): 365–368. PubMed Abstract \| Publisher Full Text\n\nKasarskis EJ, Berryman S, Vanderleest JG, et al.: Nutritional status of patients with amyotrophic lateral sclerosis: relation to the proximity of death. Am J Clin Nutr. 1996; 63(1): 130–137. PubMed Abstract\n\nReaume AG, Elliott JL, Hoffman EK, et al.: Motor neurons in Cu/Zn superoxide dismutase-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nat Genet. 1996; 13(1): 43–47. PubMed Abstract \| Publisher Full Text\n\nBruijn LI, Houseweart MK, Kato S, et al.: Aggregation and motor neuron toxicity of an ALS-linked SOD1 mutant independent from wild-type SOD1. Science. 1998; 281(5384): 1851–1854. PubMed Abstract \| Publisher Full Text\n\nKlivenyi P, Ferrante RJ, Matthews RT, et al.: Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis. Nat Med. 1999; 5(3): 347–350. PubMed Abstract \| Publisher Full Text\n\nShefner JM, Cudkowicz ME, Schoenfeld D, et al.: A clinical trial of creatine in ALS. Neurology. 2004; 63(9): 1656–1661. PubMed Abstract \| Publisher Full Text\n\nGroeneveld GJ, Veldink JH, van der Tweel I, et al.: A randomized sequential trial of creatine in amyotrophic lateral sclerosis. Ann Neurol. 2003; 53(4): 437–445. PubMed Abstract \| Publisher Full Text\n\nHedden T, Gabrieli JD: Insights into the ageing mind: a view from cognitive neuroscience. Nat Rev Neurosci. 2004; 5(2): 87–96. PubMed Abstract \| Publisher Full Text\n\nNilsson LG: Memory function in normal aging. Acta Neurol Scand Suppl. 2003; 179: 7–13. PubMed Abstract\n\nSappey-Marinier D, Calabrese G, Fein G, et al.: Effect of photic stimulation on human visual cortex lactate and phosphates using 1H and 31P magnetic resonance spectroscopy. J Cereb Blood Flow Metab. 1992; 12(4): 584–592. PubMed Abstract \| Publisher Full Text\n\nRango M, Castelli A, Scarlato G: Energetics of 3.5 s neural activation in humans: a 31P MR spectroscopy study. Magn Reson Med. 1997; 38(6): 878–883. PubMed Abstract\n\nRae C, Digney AL, McEwan SR, et al.: Oral creatine monohydrate supplementation improves brain performance: a double-blind, placebo-controlled, cross-over trial. Proc Biol Sci. 2003; 270(1529): 2147–2150. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLyoo IK, Kong SW, Sung SM, et al.: Multinuclear magnetic resonance spectroscopy of high-energy phosphate metabolites in human brain following oral supplementation of creatine-monohydrate. Psychiatry Res. 2003; 123(2): 87–100. PubMed Abstract \| Publisher Full Text\n\nMcMorris T, Mielcarz G, Harris RC, et al.: Creatine supplementation and cognitive performance in elderly individuals. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn. 2007; 14(5): 517–528. PubMed Abstract \| Publisher Full Text\n\nTiraboschi P, Hansen LA, Thal LJ, et al.: The importance of neuritic plaques and tangles to the development and evolution of AD. Neurology. 2004; 62(11): 1984–1989. PubMed Abstract \| Publisher Full Text\n\nPettegrew JW, Panchalingam K, Klunk WE, et al.: Alterations of cerebral metabolism in probable Alzheimer’s disease: a preliminary study. Neurobiol Aging. 1994; 15(1): 117–132. PubMed Abstract \| Publisher Full Text\n\nShonk T, Ross BD: Role of increased cerebral myo-inositol in the dementia of Down syndrome. Magn Reson Med. 1995; 33(6): 858–861. PubMed Abstract\n\nHuang W, Alexander GE, Daly EM, et al.: High brain myo-inositol levels in the predementia phase of Alzheimer’s disease in adults with Down’s syndrome: a 1H MRS study. Am J Psychiatry. 1999; 156(12): 1879–1886. PubMed Abstract\n\nBrewer GJ, Wallimann TW: Protective effect of the energy precursor creatine against toxicity of glutamate and beta-amyloid in rat hippocampal neurons. J Neurochem. 2000; 74(5): 1968–1978. PubMed Abstract \| Publisher Full Text\n\nCastegna A, Aksenov M, Thongboonkerd V, et al.: Proteomic identification of oxidatively modified proteins in Alzheimer’s disease brain. Part II: dihydropyrimidinase-related protein 2, alpha-enolase and heat shock cognate 71. J Neurochem. 2002; 82(6): 1524–1532. PubMed Abstract \| Publisher Full Text\n\nDavid S, Shoemaker M, Haley BE: Abnormal properties of creatine kinase in Alzheimer’s disease brain: correlation of reduced enzyme activity and active site photolabeling with aberrant cytosol-membrane partitioning. Brain Res Mol Brain Res. 1998; 54(2): 276–287. PubMed Abstract \| Publisher Full Text\n\nGallant M, Rak M, Szeghalmi A, et al.: Focally elevated creatine detected in amyloid precursor protein (APP) transgenic mice and Alzheimer disease brain tissue. J Biol Chem. 2006; 281(1): 5–8. PubMed Abstract \| Publisher Full Text\n\nTachikawa M, Fukaya M, Terasaki T, et al.: Distinct cellular expressions of creatine synthetic enzyme GAMT and creatine kinases uCK-Mi and CK-B suggest a novel neuron-glial relationship for brain energy homeostasis. Eur J Neurosci. 2004; 20(1): 144–160. PubMed Abstract \| Publisher Full Text\n\nBraissant O, Henry H, Villard AM, et al.: Creatine synthesis and transport during rat embryogenesis: spatiotemporal expression of AGAT, GAMT and CT1. BMC Dev Biol. 2005; 5: 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStraumann N, Wind A, Leuenberger T, et al.: Effects of N-linked glycosylation on the creatine transporter. Biochem J. 2006; 393(Pt 2): 459–469. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStachowiak O, Dolder M, Wallimann T, et al.: Mitochondrial creatine kinase is a prime target of peroxynitrite-induced modification and inactivation. J Biol Chem. 1998; 273(27): 16694–16699. PubMed Abstract \| Publisher Full Text\n\nStroke--1989. recommendations on stroke prevention, diagnosis, and therapy. report of the WHO task force on stroke and other cerebrovascular disorders. Stroke. 1989; 20(10): 1407–1431. PubMed Abstract \| Publisher Full Text\n\nWorld health organization: the top 10 causes of death. 2014. Reference Source\n\nHickey JV: The clinical practice of neurological and neurosurgical nursing. 5th ed. Philadelphia: Lippincott, Williams & Wilkins, 2003. Reference Source\n\nIngall T: Stroke--incidence, mortality, morbidity and risk. J Insur Med. 2004; 36(2): 143–152. PubMed Abstract\n\nWarner DS, Sheng H, Batinic-Haberle I: Oxidants, antioxidants and the ischemic brain. J Exp Biol. 2004; 207(Pt 18): 3221–3231. PubMed Abstract \| Publisher Full Text\n\nWardlaw JM, Murray V, Berge E, et al.: Thrombolysis for acute ischaemic stroke. Cochrane Database Syst Rev. 2009; (4): CD000213. PubMed Abstract \| Publisher Full Text\n\nRichard MJ, Saleh TM, El Bahh B, et al.: A novel method for inducing focal ischemia in vitro. J Neurosci Methods. 2010; 190(1): 20–27. PubMed Abstract \| Publisher Full Text\n\nYe R, Li N, Han J, et al.: Neuroprotective effects of ginsenoside Rd against oxygen-glucose deprivation in cultured hippocampal neurons. Neurosci Res. 2009; 64(3): 306–310. PubMed Abstract \| Publisher Full Text\n\nMenzies SA, Hoff JT, Betz AL: Middle cerebral artery occlusion in rats: a neurological and pathological evaluation of a reproducible model. Neurosurgery. 1992; 31(1): 100–6; discussion 106–7. PubMed Abstract\n\nMiyake K, Takeo S, Kaijihara H: Sustained decrease in brain regional blood flow after microsphere embolism in rats. Stroke. 1993; 24(3): 415–420. PubMed Abstract \| Publisher Full Text\n\nBacigaluppi M, Comi G, Hermann DM: Animal models of ischemic stroke. Part two: modeling cerebral ischemia. Open Neurol J. 2010; 4: 34–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArsava EM, Gurer G, Gursoy-Ozdemir Y, et al.: A new model of transient focal cerebral ischemia for inducing selective neuronal necrosis. Brain Res Bull. 2009; 78(4–5): 226–231. PubMed Abstract \| Publisher Full Text\n\nGideon P, Henriksen O, Sperling B, et al.: Early time course of N-acetylaspartate, creatine and phosphocreatine, and compounds containing choline in the brain after acute stroke. A proton magnetic resonance spectroscopy study. Stroke. 1992; 23(11): 1566–1572. PubMed Abstract \| Publisher Full Text\n\nCarter AJ, Muller RE, Pschorn U, et al.: Preincubation with creatine enhances levels of creatine phosphate and prevents anoxic damage in rat hippocampal slices. J Neurochem. 1995; 64(6): 2691–2699. PubMed Abstract \| Publisher Full Text\n\nBalestrino M, Rebaudo R, Lunardi G: Exogenous creatine delays anoxic depolarization and protects from hypoxic damage: dose-effect relationship. Brain Res. 1999; 816(1): 124–130. PubMed Abstract \| Publisher Full Text\n\nSestili P, Martinelli C, Bravi G, et al.: Creatine supplementation affords cytoprotection in oxidatively injured cultured mammalian cells via direct antioxidant activity. Free Radic Biol Med. 2006; 40(5): 837–849. PubMed Abstract \| Publisher Full Text\n\nWilken B, Ramirez JM, Probst I, et al.: Anoxic ATP depletion in neonatal mice brainstem is prevented by creatine supplementation. Arch Dis Child Fetal Neonatal Ed. 2000; 82(3): F224–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOtellin VA, Korzhevskii DE, Kostkin VB, et al.: The neuroprotective effect of creatine in rats with cerebral ischemia. Dokl Biol Sci. 2003; 390: 197–199. PubMed Abstract \| Publisher Full Text\n\nLensman M, Korzhevskii DE, Mourovets VO, et al.: Intracerebroventricular administration of creatine protects against damage by global cerebral ischemia in rat. Brain Res. 2006; 1114(1): 187–194. PubMed Abstract \| Publisher Full Text\n\nAdcock KH, Nedelcu J, Loenneker T, et al.: Neuroprotection of creatine supplementation in neonatal rats with transient cerebral hypoxia-ischemia. Dev Neurosci. 2002; 24(5): 382–388. PubMed Abstract \| Publisher Full Text\n\nZhu S, Li M, Figueroa BE, et al.: Prophylactic creatine administration mediates neuroprotection in cerebral ischemia in mice. J Neurosci. 2004; 24(26): 5909–5912. PubMed Abstract \| Publisher Full Text\n\nPerasso L, Spallarossa P, Gandolfo C, et al.: Therapeutic use of creatine in brain or heart ischemia: available data and future perspectives. Med Res Rev. 2013; 33(2): 336–363. PubMed Abstract \| Publisher Full Text" }	[ { "id": "6140", "date": "04 Nov 2014", "name": "James Simpkins", "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 solid contribution to the literature on creatine and its possible beneficial role in acute and chronic neurodegenerative conditions.I have one major comment and several minor issues:Major: the manuscript does not, but should, deal to any extent with possible side-effects of creatine supplementation. Particularly since the population proposed for intervention are largely elderly and very ill. A paragraph stating this and possible side-effects should suffice.Minor things:'Parkinson’s disease' section, paragraph 2, sentence 7, please insert \"not\" before \"results\". 'Amyotrophic lateral sclerosis' section, second paragraph, SOD-1 is cytosolic, so more than just mitochondrial proteins would be affected. This paragraph should be changed to reflect this. 'Long-term memory' section, third sentence, I would replace the term \"dramatic\" with another term as little drama happens in results of studies. 'Alzheimer's disease' section, first sentence - the phrase should read \"amyloid-containing plaques and neurofibrillary tangles\". 'Alzheimer's disease' section - define CRT.", "responses": [] }, { "id": "7059", "date": "16 Jan 2015", "name": "Paula Bickford", "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 review of the literature on creatine. I think that it would benefit from more discussion on possible side effects or toxicology. There are a number of review articles that discuss both the benefits and potential toxicology that could be included. Overall the article is well written and a comprehensive review of the literature. It is interesting that such diverse effects are being seen in a wide range of degenerative diseases.", "responses": [] } ]	1	https://f1000research.com/articles/3-222
https://f1000research.com/articles/3-146/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "ReactomeFIViz: the Reactome FI Cytoscape app for pathway and network-based data analysis", "authors": [ "Guanming Wu", "Eric Dawson", "Adrian Duong", "Robin Haw", "Lincoln Stein", "Eric Dawson", "Adrian Duong", "Robin Haw", "Lincoln Stein" ], "abstract": "High-throughput experiments are routinely performed in modern biological studies. However, extracting meaningful results from massive experimental data sets is a challenging task for biologists. Projecting data onto pathway and network contexts is a powerful way to unravel patterns embedded in seemingly scattered large data sets and assist knowledge discovery related to cancer and other complex diseases. We have developed a Cytoscape app called “ReactomeFIViz”, which utilizes a highly reliable gene functional interaction network and human curated pathways from Reactome and other pathway databases. This app provides a suite of features to assist biologists in performing pathway- and network-based data analysis in a biologically intuitive and user-friendly way. Biologists can use this app to uncover network and pathway patterns related to their studies, search for gene signatures from gene expression data sets, reveal pathways significantly enriched by genes in a list, and integrate multiple genomic data types into a pathway context using probabilistic graphical models. We believe our app will give researchers substantial power to analyze intrinsically noisy high-throughput experimental data to find biologically relevant information.", "keywords": [ "High-throughput experiments", "which generate large and complex data sets", "are routinely performed in modern biological and clinical studies to unravel mechanisms underlying complex diseases", "such as cancer. However", "extracting reliable and meaningful results from these experiments is usually difficult and requires sophisticated computational tools and algorithms", "which are challenging for experimental biologists to comprehend. A user-friendly software tool is extremely important for both bench and computational biologists to perform high-throughput data analysis related to cancer and other complex diseases." ], "content": "Introduction\n\nHigh-throughput experiments, which generate large and complex data sets, are routinely performed in modern biological and clinical studies to unravel mechanisms underlying complex diseases, such as cancer. However, extracting reliable and meaningful results from these experiments is usually difficult and requires sophisticated computational tools and algorithms, which are challenging for experimental biologists to comprehend. A user-friendly software tool is extremely important for both bench and computational biologists to perform high-throughput data analysis related to cancer and other complex diseases.\n\nMany studies have shown that alterations in pathways or networks are better correlated with complex disease phenotypes than any particular gene or gene product1,2. Pathway- and network-based data analysis approaches project information about seemingly unrelated genes and proteins onto pathway and network contexts, and create an integrated view for researchers to understand mechanisms related to phenotypes of interest.\n\nIn this paper, we describe a software tool called ReactomeFIViz (also called the Reactome FI Cytoscape app or ReactomeFIPlugIn), which can be used to perform pathway- and network-based data analysis for data generated from high-throughput experiments. This tool uses the highly reliable Reactome functional interaction (FI) network3 for doing network-based data analysis. The FI network was constructed by merging interactions extracted from human curated pathways with interactions predicted using a machine learning approach. This tool can also be used to perform pathway-based data analysis by using high quality human-curated pathways in the Reactome database4, the most comprehensive open source pathway database.\n\n\nImplementation\n\nWe used conventional three-tier software architecture to implement ReactomeFIViz (Figure 1). The back-end contains several databases hosted in the open-source MySQL database engine (http://www.mysql.com). The middle server-side application uses hibernate (http://hibernate.org) to access the databases storing FIs and cancer gene index data (see below). The server-side application also uses the in-house developed Reactome API for Object/Relational mapping to access pathway-related contents stored in a database using the Reactome database schema. On the server-side, a lightweight servlet container, Spring Framework (http://projects.spring.io/spring-framework/), and a Java RESTful framework, Jersey (https://jersey.java.net), are used to power a RESTful API for the Cytoscape front-end. The front-end Cytoscape app uses this RESTful API to communicate with the server-side application. Almost all analysis features in the app are provided by this RESTful API, which should also facilitate their use by other front-end applications, such as a web browser or tablet app.\n\nFor cancer data analysis, we imported the cancer gene index (CGI, https://wiki.nci.nih.gov/display/cageneindex) data into a MySQL database and then developed a hibernate API for the server-side application. The CGI data contains annotations for cancer-related genes. These annotations were extracted by using text-mining technologies and then validated by human curators (https://wiki.nci.nih.gov/display/cageneindex/Creation+of+the+Cancer+Gene+Index).\n\nThe Reactome FI network is updated annually. We recommend using the latest version of the FI network. Different versions of the FI network may yield different results due to updates to gene interactions, so we have also deployed two older versions of the FI network to use for comparison of legacy data sets and to reproduce published results. Each version is handled by its own web application on the server-side for easy software maintenance.\n\nR (http://www.r-project.org) is used for executing network module-based survival analysis and other statistical computations.\n\nReactomeFIViz implements multiple features for users to perform network-based data analysis, including FI sub-network construction, network module discovery, functional annotation, HotNet mutation analysis5,6, and network module-based gene signature discovery from microarray data sets7. By performing network-based analyses using ReactomeFIViz, researchers will be able to uncover pathway and network patterns related to their studies and then link found patterns to clinical phenotypes.\n\nThe majority of interactions in the Reactome FI network are extracted from reactions and complexes. In order to display semantic meanings (e.g. catalysis, activation and inhibition) of these interactions, we created a Reactome FI network specific visual style. This visual style is registered as a service using the OSGi API supported by Cytoscape 3, and applied to newly constructed FI sub-network automatically for network analysis.\n\nBy leveraging high quality, human annotated Reactome pathways, ReactomeFIViz allows users to explore a list of pathways, visualize Reactome pathways directly in Cytoscape, and perform pathway enrichment analysis on a list of genes. In April 2014, we added a new experimental feature for performing integrated pathway analysis for multiple genomic data types by adapting a factor graph based approach called “PARADIGM”8 into ReactomeFIViz.\n\nThe Reactome database contains several hundred manually laid-out pathway diagrams4. Pathway diagrams in Reactome are drawn based on biochemical reactions. A reaction usually contains multiple inputs and outputs, in addition to catalysts, inhibitors and activators. The network model in Cytoscape is designed to support simple graphs containing edges between two nodes only. In order to display Reactome pathway diagrams, we adapted the pathway diagram view in the Reactome curator tool4 into the Cytoscape environment, and wrapped it in a JInternalFrame so that a pathway view can be displayed along with a network view in the Cytoscape desktop (Figure 2).\n\nThe diagram view is wrapped in a JInternalFrame and hosted in the Cytoscape desktop.\n\n\nResults\n\nAs an example, we present results generated from network module based analysis for the TCGA ovarian cancer mutation data9 using ReactomeFIViz. The TCGA mutation data file and clinical information file were downloaded from the Broad Institute Firehose web site (https://confluence.broadinstitute.org/display/GDAC, released in July 2012. The clinical information has been pre-processed.). For this data set, we chose the 2009 version of the FI network, and picked genes mutated in three or more samples to construct a FI sub-network. We performed a network clustering, followed by survival analysis for each network module by splitting samples into two groups: samples having genes mutated in the module (Group 1) and samples not having genes mutated in the module (Group 0). Our results indicate that group 1 samples (Figure 3, green line in the Kaplan-Meier plot10) have significantly longer overall survival times compared to group 0 samples (Figure 3, red line in the Kaplan-Meier plot) (p-value = 3.4×10-5 based on the CoxPH analysis11). Pathway enrichment analysis results imply that module 3 is enriched with genes in calcium signaling pathway (http://www.genome.jp/kegg/pathway/hsa/hsa04020.html) and mitotic G2/M transition (http://www.reactome.org/cgi-bin/control_panel_st_id?ST_ID=REACT_2203.2). These results suggest that mutations impacting calcium signaling and the cell cycle may increase the survival of ovarian cancer patients. However, we may need more samples and independent data sets to validate our conclusion.\n\nThe central main panel shows the network view of module 3. The bottom table displays pathway annotations for genes in module 3 with two pathways, Calcium Signaling Pathway and G2/M Transition, highlighted. The right panel shows survival analysis results using both the Cox proportional hazards (CoxPH) model and the Kaplan-Meier model. The Kaplan-Meier plot was added to the figure later.\n\nUsing different versions of the FI network may yield different results because of updates of protein interactions. We performed the same analysis with the latest version of the FI network (the 2013 version), and found that genes in this module have been spread into several modules in the newer version of the FI network.\n\n\nDiscussion\n\nOur Cytoscape app provides a suite of features for users to perform network- and pathway-based analysis for data generated from multiple experiments related to cancer and other complex diseases. Users can use our tool to search for disease-related network and pathway patterns. Our tool is built upon the Reactome database, arguably the most comprehensive human curated open source pathway database, and leverages the highly reliable functional interaction network extracted from human curated pathways. Many studies based on the FI network and this app have shown its many applications to cancer and other disease studies12–15.\n\nFor future development, we will focus on using probabilistic graphical models, such as factor graphs, for performing pathway modeling and linking results to patient clinical information in order to uncover cellular mechanisms related to cancer drug sensitivity, search for cancer biomarkers, and assist new drug development.\n\n\nData availability\n\nData files used in the example: http://reactomews.oicr.on.ca:8080/caBigR3WebApp/TCGA_OV_Firehose_MAF_CLIN_2012.zip.\n\n\nSoftware availability\n\nHomepage: http://wiki.reactome.org/index.php/Reactome_FI_Cytoscape_Plugin\n\nCytoscape app: http://apps.cytoscape.org/apps/reactomefiplugin\n\nLatest source code: https://github.com/reactome-fi/CytoscapePlugIn\n\nSource code as at the time of publication: https://github.com/F1000Research/CytoscapePlugIn\n\nArchived source code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.1038516\n\nLicense: the Creative Commons Attribution 3.0 Unported License (http://www.reactome.org/?page_id=362).", "appendix": "Author contributions\n\n\n\nGW and LS initiated and guided the project. GW designed the software. GW, ED, and AD implemented the software. GW, ED, RH and LS wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis project is supported by a NIH grant (2U41HG003751-05) to LS and a Genome Canada grant (OGI 5458) to LS.\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 Irina Kalatskaya, Christina Yung and other members in Dr. Lincoln Stein’s group for software testing and many feedbacks. We also thank Peter D’Eustachio and Irina Kalatskaya for reading and editing the manuscript. We are indebted to Alexander Pico and William Longabaugh from the Cytoscape core development team for reviewing the manuscript and providing many suggestions.\n\n\nReferences\n\nBarabási 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\nVidal M, Cusick ME, Barabási AL: Interactome networks and human disease. Cell. 2011; 144(6): 986–998. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWu G, Feng X, Stein L: A human functional protein interaction network and its application to cancer data analysis. Genome Biol. 2010; 11(5): R53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCroft D, Mundo AF, Haw R, et al.: The Reactome pathway knowledgebase. Nucleic Acids Res. 2014; 42(Database issue): D472–7. PubMed Abstract \| Publisher Full Text \| Free 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\nVandin F, Clay P, Upfal E, et al.: Discovery of mutated subnetworks associated with clinical data in cancer. Pac Symp Biocomput. 2012; 55–66. PubMed Abstract \| Publisher Full Text\n\nWu G, Stein L: A network module-based method for identifying cancer prognostic signatures. Genome Biol. 2012; 13(12): R112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVaske CJ, Benz SC, Sanborn JZ, et al.: Inference of patient-specific pathway activities from multi-dimensional cancer genomics data using PARADIGM. Bioinformatics. 2010; 26(12): i237–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCancer Genome Atlas Research Network: Integrated genomic analyses of ovarian carcinoma. Nature. 2011; 474(7363): 609–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKleinbaum DG, Klein M: Survival Analysis: A Self Learning Guide. New York: Springer. 2005. Reference Source\n\nCox D: Regression Models and Life Tables (with Discussion). J R Stat Soc B. 1972; 34(2): 187–220. Reference Source\n\nShah SP, Roth A, Goya R, et al.: The clonal and mutational evolution spectrum of primary triple-negative breast cancers. Nature. 2012; 486(7403): 395–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSawey ET, Chanrion M, Cai C, et al.: Identification of a therapeutic strategy targeting amplified FGF19 in liver cancer by Oncogenomic screening. Cancer Cell. 2011; 19(3): 347–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTamborero D, Gonzalez-Perez A, Perez-Llamas C, et al.: Comprehensive identification of mutational cancer driver genes across 12 tumor types. Sci Rep. 2013; 3: 2650. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalatskaya I, Wu G: Pathway and network based analysis using Functional Interaction network. Genomics II Bacteria, Viruses and Metabolic Pathways iConcept. 2013. Reference Source\n\nWu G, Dawson E, Duong A, et al.: Reactome FI Cytoscape plugin. ZENODO. 2014. Data Source" }	[ { "id": "5320", "date": "23 Jul 2014", "name": "Hamid Bolouri", "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 describes version 4.1.0 of the Reactome Functional Interaction \"plugin\"/\"app\" for the Cytoscape network exploration/analysis Java platform. The Reactome Functional Interaction network is an established, well-received and widely used resource. The software described here provides useful resources both for exploring Reactome networks and for using them in network enrichment analyses.My only reservation about this article is that its brevity makes it sketchy. For example, how are pathway enrichment statistics calculated? Does the implementation of HotNet offered here differ in any way from the original?I am especially intrigued that the example use-case given is reported to work well with the 2009 version of the ReactomeFI network, but apparently the discovered modules cannot be found with the newer versions of the FI because the underlying interactions have been \"spread into several modules in the newer version of the FI network\". Why does this happen and how can users guard against it?I was also not able to explore/reproduce the example presented in the paper, because (contrary to its \".txt\"extension) the MAF file provided did not appear to be plain text and the needed survival data file was not included in the zip file provided. A 'README' file with instructions may be useful here.P.S. The multiple names given to this resource - \"ReactomeFIViz (also called the Reactome FI Cytoscape app or ReactomeFIPlugIn)\" - are a little confusing and seem unnecessary.", "responses": [ { "c_id": "962", "date": "06 Sep 2014", "name": "Guanming Wu", "role": "Author Response", "response": "Thanks a lot for your comments. We have made changes according to your comments. Please see details below:My only reservation about this article is that its brevity makes it sketchy. For example, how are pathway enrichment statistics calculated? Does the implementation of HotNet offered here differ in any way from the original?We have now stated that a binomial test is used for pathway enrichment analysis in the “Pathway analysis features” section and cited a reference (reference 8) for it. We also added a sentence to point out that our implementation of HotNet was done by porting the original Python and MatLab code to Java and R, and so the original algorithm is unchanged. I am especially intrigued that the example use-case given is reported to work well with the 2009 version of the ReactomeFI network, but apparently the discovered modules cannot be found with the newer versions of the FI because the underlying interactions have been \"spread into several modules in the newer version of the FI network\". Why does this happen and how can users guard against it?Because new interactions have been added into the latest version of the FI network, these new interactions may change the network clustering results. However, we still find some network modules, which overlap significantly with the original module and have significant p-values, though higher than the original one, from survival analysis. As suggested by another reviewer, we have added a supplementary document to describe results from the 2013 version of the FI network. As stated in the manuscript, we recommend that the user should use the latest version of the FI network, and check with previous versions of the FI network to investigate the stability of found network modules. I was also not able to explore/reproduce the example presented in the paper, because (contrary to its \".txt\"extension) the MAF file provided did not appear to be plain text and the needed survival data file was not included in the zip file provided. A 'README' file with instructions may be useful here.We apologize for the problem. It turned out that the zip file was collapsed somehow. We have fixed the problem, and added a simple README file as suggested. The multiple names given to this resource - \"ReactomeFIViz (also called the Reactome FI Cytoscape app or ReactomeFIPlugIn)\" - are a little confusing and seem unnecessary.Because of the development history of this software, the same app has been called different names. For example, prior to Cytoscape 3, all Cytoscape extensions were called “plug-ins”, but the nomenclature has now been changed to “app”. We have tried to minimize the confusion by using ReactomeFIViz consistently and referring to the names of earlier versions of the software just once." } ] }, { "id": "5319", "date": "11 Aug 2014", "name": "Nikolaus Schultz", "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 authors describe details of a Cytoscape 3.0 App: ReactomeFIViz. This app is intended to help users easily investigate genomic alteration data in the context of biological networks. The app allows users to obtain a network from the curated Reactome Functional Interaction database; map mutation, copy-number alteration or gene expression data onto the network; conduct a gene set enrichment analysis or module discovery on the simplified Reactome network; and finally, see the detailed pathway view provided by the Reactome Pathway Browser.The previous version of the App, which was compatible with Cytoscape 2, was introduced as a supplemental software tool in an earlier study (Wu, Feng and Stein, 2010). Since then, both Cytoscape extensions, the plug-in and the app, have actively been used by users. This has also been evident by the positive reviews the App has received on the Cytoscape App Store. Considering the rich functionality and the ease of cancer genomics analysis that the App provides to the users, we believe this app is of interest to many researchers working in the fields of Computational Biology, Cancer Genomics and Systems Biology.There is already extensive documentation about the App on the Reactome web site, however the manuscript fails to provide a general overview for non-experienced users. We have the following suggestions for the authors and if addressed, we believe, these will considerably improve the manuscript:In the abstract, the authors say “... pathways from Reactome and other pathway databases...“. The paper, however, creates the impression that the App is highly dependent on the Reactome infrastructure and does not allow communication with other databases. We suggest that the authors remove “and other pathway databases” from the abstract or better clarify this point in the abstract. The details about the type of analyses and the statistical tests the App enables are provided in an earlier paper by the same group (Wu, Feng and Stein, 2010), but not in this paper. For readers that are interested in learning these details, we suggest the authors to add a sentence to the manuscript and refer to their earlier work for details. If the functionality and the implementation of these tests have changed since then, we suggest that the authors clearly list these new improved items in the paper for more clarification. As an example use case, the authors provide the details of a re-analyzed data set and mention that the results of these two analyses differ due to changes in Reactome FI networks. Curated databases are, of course, subject to changes over time; but it is not clear from the text whether it was the changed network that was causing the problem or the new version of the extension. We suggest that the authors provide the results of these two runs as a supplement to their paper for users to compare and contrast. The last sentence of the Implementation section says that the analyses were conducted in R, but can the authors clarify the requirements for this App? Do users, for example, need to install R to use this App? Related to this, the authors do not talk about the Cytoscape version that the App targets. Is the Cytoscape 2 PlugIn deprecated? Do authors suggest that users install the newer version in Cytoscape 3 as an App? Figure 1 provides details about the implementation and the architecture of the ReactomeFIViz App, however we think the manuscript needs a simple user flow diagram that shows where different types of data are obtained, the functionalities of the App and the output the users get. The mutation-based module discovery and differential survival analysis examples mentioned in the paper are good use cases, however it is not clear from the text what the App supports other than these examples. Finally, we suggest the authors to provide a supplemental step-by-step guide to replicate the results that they describe in the paper. For a new user, this may provide a good base to start using the software and for many researchers it might be more convenient to have such an article provided as a supplementary file to this paper.", "responses": [ { "c_id": "961", "date": "06 Sep 2014", "name": "Guanming Wu", "role": "Author Response", "response": "Thanks a lot for your review of our article. We have made changes according to your thoughtful suggestions. Please see details below: In the abstract, the authors say “... pathways from Reactome and other pathway databases...“. The paper, however, creates the impression that the App is highly dependent on the Reactome infrastructure and does not allow communication with other databases. We suggest that the authors remove “and other pathway databases” from the abstract or better clarify this point in the abstract.Although the app uses a software infrastructure running on Reactome, the curated pathways it uses for the network and pathway enrichment analysis come in equal parts from Reactome and other pathway databases. We have clarified this in the abstract.The details about the type of analyses and the statistical tests the App enables are provided in an earlier paper by the same group (Wu, Feng and Stein, 2010), but not in this paper. For readers that are interested in learning these details, we suggest the authors to add a sentence to the manuscript and refer to their earlier work for details. If the functionality and the implementation of these tests have changed since then, we suggest that the authors clearly list these new improved items in the paper for more clarification.We have added references to our original FI network paper in the paragraph introducing “Network analysis features”. Also, we added the date that version 4.0.0.beta was released in order to indicate that pathway analysis features are new features that have not been covered by our previous work. As an example use case, the authors provide the details of a re-analyzed data set and mention that the results of these two analyses differ due to changes in Reactome FI networks. Curated databases are, of course, subject to changes over time; but it is not clear from the text whether it was the changed network that was causing the problem or the new version of the extension. We suggest that the authors provide the results of these two runs as a supplement to their paper for users to compare and contrast.We modified the sentence in the last paragraph describing different results using different versions of the FI network to make it clear that the same software, but different versions of the FI network, were used. We also deleted this sentence in the implementation section, “Each version is handled by its own web application on the server-side for easy software maintenance.”, to avoid confusing the readers with technical details.As suggested, added a supplementary document to describe results from the 2013 version of the FI network. The last sentence of the Implementation section says that the analyses were conducted in R, but can the authors clarify the requirements for this App? Do users, for example, need to install R to use this App? Related to this, the authors do not talk about the Cytoscape version that the App targets. Is the Cytoscape 2 PlugIn deprecated? Do authors suggest that users install the newer version in Cytoscape 3 as an App?We have clarified the section to make the requirements clearer, and added a new paragraph to address the relationship of the application to Cytoscape 2 and 3. Figure 1 provides details about the implementation and the architecture of the ReactomeFIViz App, however we think the manuscript needs a simple user flow diagram that shows where different types of data are obtained, the functionalities of the App and the output the users get. The mutation-based module discovery and differential survival analysis examples mentioned in the paper are good use cases, however it is not clear from the text what the App supports other than these examples.We have added a new paragraph at the top of the Results section to highlight some major features in the App, and created a new diagram to show these major features as suggested. Finally, we suggest the authors to provide a supplemental step-by-step guide to replicate the results that they describe in the paper. For a new user, this may provide a good base to start using the software and for many researchers it might be more convenient to have such an article provided as a supplementary file to this paper.The application’s web-based tutorial already has provided very detailed, step-by-step instructions that replicate the results. We have added a new sentence in the Data availability section to point this out, and the README file contained in the zip file for downloading now points readers interested in replicating the analysis to the online tutorial." } ] } ]	1	https://f1000research.com/articles/3-146
https://f1000research.com/articles/3-6/v1	09 Jan 14	{ "type": "Web Tool", "title": "DataUp: A tool to help researchers describe and share tabular data", "authors": [ "Carly Strasser", "John Kunze", "Stephen Abrams", "Patricia Cruse", "John Kunze", "Stephen Abrams", "Patricia Cruse" ], "abstract": "Scientific datasets have immeasurable value, but they lose their value over time without proper documentation, long-term storage, and easy discovery and access. Across disciplines as diverse as astronomy, demography, archeology, and ecology, large numbers of small heterogeneous datasets (i.e., the long tail of data) are especially at risk unless they are properly documented, saved, and shared. One unifying factor for many of these at-risk datasets is that they reside in spreadsheets.In response to this need, the California Digital Library (CDL) partnered with Microsoft Research Connections and the Gordon and Betty Moore Foundation to create the DataUp data management tool for Microsoft Excel. Many researchers creating these small, heterogeneous datasets use Excel at some point in their data collection and analysis workflow, so we were interested in developing a data management tool that fits easily into those work flows and minimizes the learning curve for researchers.The DataUp project began in August 2011. We first formally assessed the needs of researchers by conducting surveys and interviews of our target research groups: earth, environmental, and ecological scientists. We found that, on average, researchers had very poor data management practices, were not aware of data centers or metadata standards, and did not understand the benefits of data management or sharing. Based on our survey results, we composed a list of desirable components and requirements and solicited feedback from the community to prioritize potential features of the DataUp tool. These requirements were then relayed to the software developers, and DataUp was successfully launched in October 2012.", "keywords": [ "The move towards digital data is ubiquitous across all domains in academic research and scholarship1–5", "and these data can be made available more easily and distributed more quickly than ever before. This is often called the data deluge", "and is a phenomenon that has been examined in the traditional academic literature2", "4", "6", "as well as in several major media outlets7–9." ], "content": "Introduction\n\nThe move towards digital data is ubiquitous across all domains in academic research and scholarship1–5, and these data can be made available more easily and distributed more quickly than ever before. This is often called the data deluge, and is a phenomenon that has been examined in the traditional academic literature2,4,6, as well as in several major media outlets7–9.\n\nAmong the most pressing problems associated with the data deluge is good data management: how does one handle the huge volume of available information effectively and efficiently to solve important problems? Knowledge of good data management techniques and software development lags behind the progression of the data deluge. Consequently, although researchers of all fields are faced with huge volumes of data from increasingly diverse sources, they do not have the skills to handle their data sets. This challenge is amplified by the fact that research data are seldom shared, re-used, or preserved10–12. There is a growing awareness among practitioners and funders that this situation represents inefficient use of research dollars, missed opportunities to exploit prior investment, and a general loss for the scholarly community13. Michener et al.14 described the loss of valuable data and insight about those datasets as “information entropy”. This loss of information is becoming increasingly worrisome as data management practices improve very slowly, while the volume of data grows exponentially.\n\nRecognizing that most earth, environmental, and ecological scientists use spreadsheets at some point in their data life cycle, the California Digital Library (CDL) partnered with Microsoft Research Connections and the Gordon and Betty Moore Foundation to create a tool that would encourage and enable good data stewardship practices for datasets created in Microsoft Excel. Our vision was to promote publishing, archiving, and sharing of tabular data among earth, environmental, oceanographic, and ecological scientists by creating a tool that will easily integrate into their current workflows and assist them in data management and preservation. This will, in turn, enable faster and more efficient research, thereby increasing the pace of scientific advancement.\n\nTo optimize the tool, we first identified the needs of the community via surveys of researchers. We found that, on average, researchers had poor data management practices, were not aware of data centers or metadata standards, and did not understand the benefits of data management or sharing. We used the survey results to compose a list of desirable components and solicited feedback from the community to prioritize potential features.\n\nThe resulting DataUp tool facilitates documenting, managing, archiving, and sharing tabular scientific data. It comes in two forms, both open-source: an add-in for Excel and a web-based application. The add-in operates within the well-known program Microsoft Excel; the web application allows users to upload tabular data to the web-based tool in either Excel (.xlsx) or comma-separated value (.csv) format. Both the add-in and the web application provide users with the ability to (1) Perform a “best practices check” to ensure the data are CSV-compatible; (2) Create standardized metadata, or a description of the data, using a wizard-style template; (3) Retrieve a unique identifier for their dataset from their chosen data repository, and (4) Post their datasets and associated metadata to the repository.\n\n\nMethods and results\n\nThe extent to which researchers use Microsoft Excel is not fully documented, however based on strong anecdotal evidence we assumed that it is a standard tool for most scientists. Given this fact, we determined that an add-in for Excel would have the greatest potential impact on how scientists work with data. An add-in (also called a plug-in) is a small piece of software that one installs on a local computer. Once installed, it extends the capabilities of an existing program: in this case, Excel. The add-in’s functionality is available from within the program, and in the case of Excel, appears as a “ribbon” of functions and features within the standard user interface. In this way, we assumed that researchers would be more likely to use the tool since it is fully integrated with a program they are already using.\n\nOur target audience for creating the tool was scientists and researchers actively working with earth, environmental, oceanographic, and ecological data. These researcher groups were chosen based on their relatively low participation in data sharing15 and their presumed high levels of Excel use. To capture their data management needs, we surveyed and interviewed more than 130 researchers over the course of five months (August to December, 2011). We also collected suggestions for requirements from academic libraries, data centers, data managers, and other data professionals, although this collection was less structured and more anecdotal. Most of these interactions occurred via interactions with the DataONE project community16; a full list of partners affiliated with the DataONE project is available on their website (http://dataone.org).\n\nWe used several methods to communicate with our potential stakeholder community in developing the tool. These included the DCXL blog (now the Data Pub Blog, located at datapub.cdlib.org), two Twitter accounts (@dataupcdl and @carlystrasser), and interviews and conversations at conferences, webinars, and professional meetings.\n\nOur goal in surveying and interviewing researchers was to determine how they were currently handling data management, especially as it related to Excel data, and how best the tool we were developing might help improve researcher practices surrounding data. The questions we asked underwent revision to improve the survey instrument, and to that end we used four similar versions of the survey over the course of data collection. The number of respondents for each survey version was 43, 12, 47, and 10 respectively, for a total of 112 respondents. The four versions of the survey can be viewed in the associated datasets. Interview questions were less structured and varied depending on the interviewee.\n\nWe attended four professional meetings and surveyed researchers of various statuses (i.e., from student to senior researcher) and from many different institutions and organizations (Table 1). We also conducted surveys and in-depth interviews with researchers at four campuses in the University of California system from September 2011 to February 2012. Interviewees volunteered to participate by contacting one of the authors, Carly Strasser, directly. Overall, we collected 112 surveys and conducted 30 interviews (of 30 to 90 minute duration) from 133 people representing 84 different institutions (Table 1). Less formal information was obtained from other venues, including comments on the DataUp Blog, discussions with librarians and data center managers, and conversations with researchers at DataONE meetings.\n\nDemographically, the survey pool was composed of researchers and scientists ranging from undergraduate-level to PhD-level (Figure 1).\n\nn = 133.\n\nWe asked researchers about their choice of operating system because of the potential implications for development of the tool. Of those surveyed, the large majority (74%) used a Windows-based operating system, while 23% used a Mac-based system and 2% used Linux (Figure 2).\n\nn = 133.\n\nWe asked a series of questions related to how the researchers were using Excel for their day-to-day work. We found that 80% of those surveyed answered that they used Excel “every day” or “almost every day” (Figure 3).\n\nn = 118.\n\nWhen asked what data-related tasks they were undertaking when using Excel, we found that most were at least using Excel to organize their data (96%). Excel was also used by the majority of participants for visualizing data (61%), performing minor calculations (75%), and for sharing data with colleagues in Excel format (81%) (Figure 4).\n\nn = 119.\n\nTo better understand the content of researchers’ spreadsheets, we asked whether the following Excel features were used in their datasets (Figure 5).\n\nn = 70.\n\nmultiple tables on a single spreadsheet\n\nmultiple tabs within an Excel file\n\nheader row with parameter labels created\n\nunits provided alongside data (i.e., in the data cell or header row)\n\nembedded formulas\n\npivot tables\n\nmacros\n\nembedded comments\n\ncell shading to indicate information about the data (i.e., ad-hoc metadata)\n\nMost researchers created header rows (97%), used embedded formulas (83%), and used cell shading as a form of ad-hoc metadata (74%). Of those we surveyed, the majority (74%) reported that they had a “better than average” knowledge of the Excel software, while 24% reported an average knowledge (n = 105).\n\nWe asked researchers to identify other software programs that they use alongside Excel for their data analysis and organization. Note that these results are likely heavily influenced by the venues used to interview researchers, since software programs tend to be used by many researchers in a given discipline (Figure 6). The open-source statistical software R was most often cited (53%).\n\nn = 131.\n\nOther information gathered via the survey included areas of work (i.e., field versus lab; area of focus; discipline), attitudes about data sharing, and knowledge of data repositories. These questions were not asked formally via survey in most cases, rendering the results difficult to share with any confidence in the numbers reported.\n\nAlthough the practices reported by researchers are common and accepted uses of Excel, they are not necessarily well suited for long-term preservation of high-quality data. This has been previously reported in the literature17–20. In addition, the European Spreadsheet Risk Interest Group has created a curated list of stories detailing instances where spreadsheets are implicated in erroneous reporting (http://eusprig.org). In general, issues associated with using Excel for generating curation-ready datasets are (1) poor data table construction (e.g. multiple data tables on a single spreadsheet); (2) a lack of metadata or poorly standardized metadata (e.g. using comments, notes, color-coding, and shading to document important details about the dataset; (3) embedded figures, charts, and comments that make the spreadsheet less usable in programs outside Excel; and (4) poor provenance of how data is produced via calculations, statistics, and formulas.\n\nBased on the information collected from researchers and other stakeholders, we created the following high-level requirements for the tool:\n\n1. Check data file for .csv compatibility and create .csv version data file. The user can generate and download a customized report detailing elements in their dataset that might cause problems for data archiving and/or export of the data file as a .csv version.\n\n2. Generate metadata that is linked to the data file. Using the DataUp tool, machine- and human-readable metadata is generated, embedded in the data file, and can be exported as a separate file. The metadata is displayed in a new tab on the spreadsheet, can be saved separately, and relies on Ecological Metadata Language (EML) and the DataONE metadata schema (http://mule1.dataone.org/ArchitectureDocscurrent/design/SearchMetadata.html). Both file-level and parameter-level metadata are created by the tool.\n\nFile-level metadata is information about the entire dataset, such as the creator, temporal and spatial details of the data collection, and the funders of the project. The tool is able to pre-populate some fields based on user information provided by Excel. Keywords can be selected from standard lists.\n\nParameter metadata describes individual elements of the data file, and most commonly corresponds to the header row of a tabular dataset. The user can identify a header row to begin the process of creating parameter metadata.\n\n3. Generate a citation for the data file. Using the tool, the user can generate a complete data citation for their tabular dataset. This includes all the metadata necessary for citing the dataset, is in a standard format, and becomes part of the metadata. The citation can be downloaded in standard formats (e.g. .ris, .bib, .xml).\n\n4. Repository authentication set-up. The user can authenticate with their chosen repository from within the tool, assuming they have pre-existing login information for that repository. This will then allow them to deposit their dataset in the repository via the tool.\n\n5. Link an identifier to the data file. The tool allows the user to retrieve and save a persistent identifier (such as a DOI) for their dataset from their chosen repository.\n\n6. Ensure that the data file is ready for deposition into a repository. The tool determines whether the data file is ready for deposit into the designated archive by checking for the following:\n\nDetermine whether a compatibility check has been completed.\n\nDetermine whether metadata is complete (i.e., all required metadata are present).\n\nDetermine whether a citation has been generated.\n\nThe tool then generates the technical metadata needed by the designated repository.\n\n7. Submit the data file for deposition into the designated repository.\n\n8. Ensure compatibility for Excel users without the add-in: users without the add-in locally installed are able to open the data file and access the metadata.\n\nThese requirements were posted on the DataUp blog, with requests for feedback from the community. We then passed on the document to the Microsoft Research team, who generated a second version of the requirements based on their knowledge of Excel and their protocols for software development. These requirements were then relayed to the developers (contractors for Microsoft Research).\n\nIn the course of development, questions arose from the project team as to whether an Excel add-in was the most appropriate choice for delivering the tool to researchers; the alternative discussed was a web-based application. Concerns were that an add-in had compatibility issues that required updates on the developer’s part and downloads on the user’s part. In addition, the project timeline dictated that the add-in could be built only for Windows platforms; Macintosh systems would not be able to use the tool. This is not true for a web-based application. See Table 2 for a summary of the differences between the two potential versions of the tool: an add-in and a web application.\n\nIn early 2012, we launched a campaign to determine which of the two versions of the tool should be created. Input was received from attendees of the Ocean Sciences 2012 Meeting in Salt Lake City, Utah. We also asked researchers and others via online surveys and blog posts which they would prefer, and what barriers they perceived to each version of the tool. We collected results from approximately 200 individuals. Most (95%) were willing to download an add-in, and most (83%) indicated that they would prefer an add-in to a web application (assuming the add-in were available for Mac as well). However 72% reported that there were barriers to their downloading and/or installing an add-in for Excel. Barriers mentioned included version compatibility issues, security concerns (e.g., viruses), lack of Mac compatibility, and a lack of administrative controls over computers, preventing downloads. The full set of survey responses is available in the associated datasets.\n\nGiven these contradictory results we determined that there was a need for both versions of the tool. We therefore proceeded with the development of both an add-in for Excel and a web-based application. The requirements were the same for both versions; only the delivery of the functionality differed between the two. Of those surveyed, 75% used a Windows operating system, compared to 22% using a Mac, and 3% using some other system (e.g. Linux). These results paralleled those from our general researcher survey (Figure 2).\n\nThe tool created based on our requirements and user feedback is called DataUp. DataUp is free and open source, and has two forms: a web-based application (web app http://dataup.org) and a downloadable Excel add-in. Both versions of the tool provide users with the ability to (1) perform a “best practices check” to ensure that data are well formatted and organized; (2) create standardized metadata (i.e., a scientifically-meaningful description of the data), using a wizard-style template; (3) retrieve a unique identifier for their dataset from their chosen data repository; and (4) upload datasets and associated metadata to a public data repository.\n\nBest practices check. The tool determines whether the data file has any of 11 potential issues that do not comply with data management best practices, such as embedded charts, comments, and color-coded cells. These issues were chosen based on interviews with researchers, as well as data managers who often “clean up” spreadsheets submitted by researchers for archiving. In addition to identifying the locations of these problems, DataUp informs the user why they are potentially problematic, and offers suggested alternatives or the ability to remove them in bulk. The information provided by the DataUp tool for each of these potential issues is below:\n\n1. Embedded charts, tables, pictures. Why: These embedded items will not be visible when data are exported as a .csv file. Also, these elements are visible only if the file is opened with Microsoft Excel. Suggested remedy: Move embedded charts, tables, or pictures to other tabs in your file or to a completely separate file.\n\n2. Embedded comments. Why: Comments will not be visible when data are exported as a .csv file. Also, these elements are visible only if the file is opened with Microsoft Excel. Suggested remedy: Create a new column titled “Comments” and add your text there.\n\n3. Commas. Why: Commas are often used to separate multiple piece of information/data (e.g. City, State). Cells only contain one piece of information. Suggested remedy: Split pieces of information into multiple columns (e.g. City column and State column).\n\n4. Special characters. Why: Special characters may cause problems for other programs or may be modified upon export. Suggested remedy: Use alpha-numeric characters only. If needed, describe the symbol in a new column.\n\n5. Color coded text or cell shading. Why: Formatting will not be visible when data are exported as a .csv file. If formatting is used as a coding scheme, all codes will be lost upon export. Suggested remedy: Use descriptions or alphanumeric coding schemes in a new column.\n\n6. Columns have mixed data types. Why: Some programs cannot handle mixed data types (e.g. numbers and text in the same column). Suggested remedy: Ensure you are using only numbers or only text in a column; split data into multiple columns if necessary.\n\n7. Non-contiguous data. Why: Empty columns or rows tend to be used to separate multiple data tables on the same tab. Suggested remedy: Move multiple tables onto separate tabs.\n\n8. Merged cells. Why: Merged cells will not be maintained when data are exported as a .csv file. Information may be lost when cells are un-merged upon export. Suggested remedy: Un-merge cells and annotate appropriately so information is not lost.\n\n9. Blank cells. Why: Blank cells within a contiguous data table are potentially problematic for reading files in other programs. Suggested remedy: Designate a coding scheme for missing data or other explanations for blank cells.\n\n10. Header row absent or more than one header row. Why: Ideally the first row of a data table contains parameter names for the columns. If there is no header row, your data table may be difficult to use and document. If there are multiple header rows, some software programs may have problems. Suggested remedy: Create a header row with unique parameter names that describe the column’s contents.\n\n11. Multiple sheets (tabs). Why: Multiple sheets will not be maintained as a single document if the file is converted to .csv. Suggested remedy: The user can move each tab into a separate .csv file. If left as multiple sheet,s the DataUp tool will automatically export the data as separate .csv files.\n\nCreate metadata. DataUp helps the researcher create standard metadata using a form that becomes part of their spreadsheet, facilitating future use and sharing. Metadata can be generated at both the file- and column-level. File-level metadata includes names, email addresses and institutional affiliations for project personnel, and dataset titles. Column-level metadata (i.e. attribute metadata) includes information about the variables in the dataset, the units of measure, and descriptions of each column of data. DataUp creates metadata using the Ecological Metadata Language (EML). This particular standard was chosen because of its widespread use in our original target communities. In addition, EML is both flexible and extensible, which enables future modifications to the chosen schema as necessary. We selected 47 elements of EML for DataUp, with seven elements required (Table 3). We choose to support only a subset of EML in order to provide the lowest barrier to entry for researchers interested in documenting their datasets.\n\n*elements are required.\n\nObtain an identifier. Valuing and incentivizing the time and effort required to manage data well is an important factor in fostering data sharing and reuse. One way to allow data producers to get credit for this is through data citation. The DataUp tool connects to the user’s chosen repository to retrieve a unique identifier for the researcher’s dataset. For its first iteration, DataUp connects to the EZID service (http://n2t.net/ezid), based at CDL, used by the public DataUp ONEShare repository. The identifier generated is an ARK (Archival Resource Key, https://confluence.ucop.edu/display/Curation/ARK). ARKs provide stable, opaque, versatile, and transcription-safe identifiers. This identifier is saved in the data file’s metadata.\n\nShare and archive. Once metadata is created, the user can connect directly to a repository via DataUp and upload their data for archiving. Currently, DataUp is connected to ONEShare, which is a dedicated public DataUp repository to which anyone may deposit tabular data (more information below).\n\n\nArchitecture\n\nDataUp’s codebase is written in C# using the .NET application framework. The web app is deployed on Microsoft’s Windows Azure cloud platform. DataUp’s architecture (Figure 7) consists of two clients communicating via an intermediating web service to one or more repositories. The add-in client is an Excel extension that runs directly on a researcher’s Windows-based computer. The web app client runs as an online application hosted in Azure. Client/web service communication uses the OData protocol 21 . Both clients support standard EML metadata and draw functionality from a common web service, also hosted in Azure. That web service is managed by a separate administrative service.\n\nDataUp was designed not only for standalone metadata checks, but also for contacting a variety of repositories to obtain persistent identifiers and to archive data. Currently, the only repository supported is ONEShare, an instance of the CDL Merritt repository that is also a DataONE Member Node (more information below). With the front-end running at CDL and a storage node back-end running at the University of New Mexico, content can be browsed either by logging in directly to Merritt as a guest or using the DataONE ONEMercury interface (http://dataone.org/onemercury).\n\nAlthough there are hundreds of data repositories available to researchers for data archiving, the majority of scientists are not aware of their existence or how to access them. One of the major outcomes of the DataUp project is the ONEShare repository, created specifically for the DataUp tool. ONEShare is a special instance of CDL’s Merritt repository, which serves as a digital archive and access system to the University of California campuses (http://www.cdlib.org/uc3/merritt). Users can deposit their tabular data and metadata directly into the ONEShare repository from within the tool, allowing for seamless data archiving within the researcher’s current workflow. The DataUp web service performs the repository submission using the Merritt API, hiding all details of the transfer protocol from the DataUp user. An added advantage of ONEShare is its connection to the DataONE network of repositories. DataONE links together existing data centers and enables its users to search for data across all participating repositories using a single search interface. Since Merritt is a member node on the DataONE network, all data deposited into ONEShare will be indexed and made discoverable by any DataONE user, facilitating collaboration and enabling data re-use.\n\nThe ONEShare repository is collaboratively supported by the CDL and the University of New Mexico Library. CDL’s Merritt repository relies on a highly decentralized micro-services architecture22. In the case of ONEShare, a Merritt storage node was established on a University of New Mexico (UNM) virtual server managing a local file system. All DataUp submissions to Merritt are routed automatically to the UNM storage node, but the data are still subject to all Merritt preservation and access services such as ongoing fixity audits, metadata search and browse, and pro-active preservation analysis and planning. Merritt is also integrated as a member node on the DataONE network and the full set of descriptive metadata for all DataUp-submitted data is automatically harvested by the DataONE coordinating nodes for inclusion in the federated ONEMercury search interface, increasing the public visibility of DataUp datasets.\n\nThe first versions of the add-in and web application underwent beta testing by researchers, librarians, software engineers, and other stakeholders. Testers included professional contacts of the DataUp team, researchers who participated in the requirements-gathering survey and consented to future contact, and individuals responding to a blog post requesting subjects for beta testing. We received feedback from 23 testers via an online survey. We received additional comments via email and conversations with researchers. Information gathered from the beta testers was relayed to the developers who addressed those issues that were reasonable within the given time frame for software release. Data from beta testing is available from the associated datasets.\n\nOne of the major goals of the DataUp project was to create an open-source tool that could be adopted and used by the larger community. To that end, we partnered with the non-profit Outercurve Foundation, whose goal is to enable code exchange and understanding among software companies and open source communities. The DataUp project site for Outercurve holds the copyright to DataUp code, and has released it under Apache2.0 license (https://www.outercurve.org/Galleries/ResearchAccelerators/DataUp). The code for all aspects of the DataUp tool (add-in, web app, and web service) is available on the project’s BitBucket site (http://www.bitbucket.org/DataUp). Minimum system requirements for the web application are an internet connection and web browser. For the add-in, the user must be running a Windows operating system with Microsoft Excel 2007 or higher.\n\n\nDiscussion and conclusions\n\nResponse to the release of the tool was enthusiastic. Between October 2012 and December 2013, the add-in version of the tool had been downloaded more than 700 times, and we estimate a proportionate interest in the web app version of the tool. The main DataUp website has had over 17,000 page views with visitors from more than 10 countries (84% of visits from the US). These numbers do not, however, adequately represent the tool’s popularity and potential. The CDL has received inquiries about DataUp from many repositories, organizations, and publishers interested in configuring the tool for their needs. The inquiries represent a range of stakeholders that are crucial to data sharing, including a large citizen science project, a major social science data archive, some highprofile data publication services, and others. They are excited about the possibilities that DataUp represents for linking researchers’ workflows directly to repositories, with capabilities for generating metadata and performing best practices checks.\n\nThe DataUp team received a one-year grant from the US National Science Foundation, supplemental to the DataONE project. Using these funds, the DataUp web application will undergo another iteration that will result in easier repository connections, better features, and a more streamlined workflow. The code resulting from this project will be open-source, and community ownership will be encouraged. The text of the NSF proposal is available from the University of California’s eScholarship repository23.\n\nCDL envisions that the future of DataUp will be directed by the community of stakeholders. Interested developers can expand upon and increase the tool’s functionality to meet the needs of a broad array of researchers. Code for both the add-in and web application is open source and participation in its improvement is strongly encouraged. Although the target audience for our tools that result from the DataUp project will be earth, environmental, oceanographic, and ecological scientists, we envisage that any tools developed will be easily implemented in other research communities, such as the social sciences.\n\n\nData and software availability\n\nFigshare: DataUp manuscript data, doi: 10.6084/m9.figshare.88462524.\n\nZenodo: The DataUp source code package, doi: 10.5281/zenodo.763925.\n\nBitbucket: Source code for the DataUp Excel add-in and web application, https://bitbucket.org/dataup/.", "appendix": "Author contributions\n\n\n\nC.S. was the DataUp project manager and the primary author on the manuscript. J.K. and S.A. helped with technical details and worked on technical aspects of the project. P.C. provided oversight for development of DataUp. All authors contributed to the writing and editing of this manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a grant to the California Digital Library from the Gordon and Betty Moore Foundation (Grant No 2736) and from the Microsoft Research (MicrosoftPO 95341291 ESC; UMS 18615; Extension: 95343632).\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 acknowledge the support of team members from Microsoft Research (L. Dirks, K. Tolle). We also thank C. Mentzel (GBMF) for his support and insight during the project. Documentation for the tool was written in part by I. Rabinovitch and S. Sinick. Developers at the CDL who contributed to the project’s successful launch were S. Fisher, D. Loy, M. Reyes, and P. Willett. Members of the DataONE project team helped with survey distribution and beta testing; these members include (but are not limited to) D. Vieglais, A. Budden, W. Michener, R. Koskela, and M. Jones. CDL staff that contributed heavily to the project’s success include S. Lew, B. Yung, and R. Epting-Day. We also thank Outercurve’s E. Schultz and colleagues for their help in making the DataUp project open-source. Special thanks to K. Ram for insight during the project’s development and execution.\n\n\nReferences\n\nInteragency Working Group on Digital Data. Harnessing the power of digital data for science and society. 2009. Reference Source\n\nCarlson S: Lost in a sea of science data. The Chronicle of Higher Education. 2006; 52(42): A35. Reference Source\n\nBorgman C: The digital future is now: A call to action for the humanities. Digital Humanities Quarterly. 2009; 3(4). Reference Source\n\nFaniel I, Zimmerman A: Beyond the data deluge: A research agenda for large-scale data sharing and reuse. The International Journal of Digital Curation. 2011; 6(1): 58–69. Publisher Full Text\n\nBorgman C: Data, disciplines, and scholarly publishing. Learned Publishing. 2008; 21: 29–38. Publisher Full Text\n\nBorgman C, Wallis J, Enyedy N: Little science confronts the data deluge: habitat ecology, embedded sensor networks, and digital libraries. International Journal on Digital Libraries. 2007; 7(1–2): 17–30. Publisher Full Text\n\nThe Economist Editors. The data deluge: Business, governments and society are only starting to tap its vast potential. The Economist. 2010. Reference Source\n\nPollack A: DNA Sequencing Caught in the Deluge of Data. New York Times. 2011. Reference Source\n\nBell G, Hey T, Szalay A: Computer science. Beyond the data deluge. Science. 2009; 323(5919): 1297–1298. PubMed Abstract \| Publisher Full Text\n\nNelson B: Data sharing: Empty archives. Nature. 2009; 461(7261): 160–163. PubMed Abstract \| Publisher Full Text\n\nTenopir C, Allard S, Douglass K, et al.: Data sharing by scientists: Practices and perceptions. PLoS One. 2011; 6(6): e21101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeClere F: Too many researchers are reluctant to share their data. 2010. Reference Source\n\nNature Editors. Data’s shameful neglect. Nature. 2009; 461(7261): 145. PubMed Abstract \| Publisher Full Text\n\nMichener W, Brunt J, Helly J, et al.: Nongeospatial metadata for the ecological sciences. Ecol Appl. 1997; 7(1): 330–342. Publisher Full Text\n\nHampton S, Strasser C, Tewksbury J: Growing pains for ecology in the twenty-first century. BioScience. 2013; 63(2): 69–71. Publisher Full Text\n\nMichener W, et al.: DataNetONE: Observation Network for Earth. (NSF Grant No. OCI 0830944), 2009. Publisher Full Text\n\nWolstencroft K, Owen S, Horridge M, et al.: RightField: Embedding ontology annotation in spreadsheets. Bioinformatics. 2011; 27(14): 2021–2022. PubMed Abstract \| Publisher Full Text\n\nMichener W, Beach J, Jones M, et al.: A knowledge environment for the biodiversity and ecological sciences. J Intell Inf Syst. 2007; 29(1): 111–126. Publisher Full Text\n\nLinden J, Green A: Don’t leave the data in the dark. D-Lib Magazine. 2006; 12(1): 48–57. Publisher Full Text\n\nLeong K: The seven deadly spreadsheet sins. Blog, Production-Scheduling.com. 2013. Reference Source\n\nMicrosoft Corporation. [MS-ODATA]: Open Data Protocol (OData). 2013. Reference Source\n\nAbrams S, Cruse P, Kunze J, et al.: Curation micro-services: A pipeline metaphor for repositories. J Digit Imaging. 2011; 12(2). Publisher Full Text\n\nDataUp: Further Development and Community Building. eScholarship. 2013. Reference Source\n\nStrasser C, Cruse P, Kunze J, et al.: DataUp manuscript data. Figshare. 2014. Data Source\n\nStrasser C, Cruse P, Kunze J, et al.: The DataUp source code package. ZENODO 2014. Data Source" }	[ { "id": "4573", "date": "02 May 2014", "name": "Carol Goble", "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 the DataUp Excel-based metadata and data capturing tool developed as part of the DataONE project. I like the paper and I like the tool.The paper is well written. The need for spreadsheet-based data management tools is critical, and DataUp makes a valuable contribution. The tool works, the software is available, is being used, and is useful. The survey results and the requirements are a very useful guide for other workers using spreadsheets as a prime mechanism for data upload. The survey is well conducted given the constraints of such things, and it is refreshing to see that the people who do the data management (postdocs, postgrads) were targeted. This is a very useful contribution to the field.There are, however, some improvements that I would like to see in the final article:The metadata model seems to be very high level (Table 3). Is there a richer metadata model for specific data types? Is it possible to upload and index/search on more domain specific metadata captured in the DataUp model? To what extent does the metadata model work for all the data types you mention? As your user base is wide one would expect heterogeneity to be a big problem. How are controlled vocabularies and or specific domain metadata models incorporated? The architecture figure 7 is a very general figure and can be replaced by something that showed the protocol of how DataUp is used in practice. Are Excel templates prepared by the DataUp team or through the Plug-in? There is sparse information on uptake or the impact of uptake. The number of downloads are listed but not how many datasets were uploaded to the repository using DataUp. Despite the excellent requirements survey and user engagement, there is no evaluation of DataUp’s use. What is the difference in uptake between the Excel and web-based version? There is no related work section. The only related work is RightField (reference17) but what RightField does and how it relates to DataUp is not mentioned. Similar tools to DataUp such as ISAtool Suite and Ontomaton are not mentioned. The survey appears USA-centric. EZID is only available in the USA. DataUp currently works with DataONEShare. Can and how DataUp be adapted for use in other repositories? Can it reuse infrastructure that is not US based? Very minor comments:There are two typos:sheet,s -> sheetshighprofile -> high profile", "responses": [ { "c_id": "946", "date": "18 Aug 2014", "name": "Carly Strasser", "role": "Author Response", "response": "Reviewer text is italicized; our responses are below the reviewer text.The metadata model seems to be very high level (Table 3). Is there a richer metadata model for specific data types? Is it possible to upload and index/search on more domain specific metadata captured in the DataUp model? To what extent does the metadata model work for all the data types you mention? As your user base is wide one would expect heterogeneity to be a big problem.Currently there is not a richer metadata model. Future development plans include the ability to expand metadata options and allow for generating metadata files that comply with different standards for various disciplines. We chose some elements of EML as our baseline for three reasons: (1) the metadata fields are fairly generic, so they can apply to many different disciplines; (2) EML is a flexible metadata language that, although originally constructed for Ecology, has the ability to describe a wide variety of datasets; and (3) EML is one of the metadata standards accepted by the DataONE network.How are controlled vocabularies and or specific domain metadata models incorporated? The architecture figure 7 is a very general figure and can be replaced by something that showed the protocol of how DataUp is used in practice. Are Excel templates prepared by the DataUp team or through the Plug-in?There are no controlled vocabularies incorporated into the tool, although this has been on our “wish list” of future development. Currently a user can specify a controlled vocabulary in the metadata, however we have no connection or integration with vocabulary systems (and no current mechanism for that connection). Similarly, the metadata must be hard-coded and therefore we have no ability to “switch out” the metadata depending on domain-specific interests.There are no templates in use by the DataUp tool. Instead, there are a set of “rules” that the tool consults while parsing a user’s spreadsheet. The tool checks for best practices and reports back; this is not in any way set by the DataUp team. There is sparse information on uptake or the impact of uptake. The number of downloads are listed but not how many datasets were uploaded to the repository using DataUp. Despite the excellent requirements survey and user engagement, there is no evaluation of DataUp’s use. What is the difference in uptake between the Excel and web-based version?Unfortunately we don’t have access to this information, nor do we believe it is being collected by the service. We are limited in our ability to modify the code base to obtain these metrics since our technical team is not familiar with the Microsoft Azure service or the C#/.NET code base. We can cite the number of downloads of the add-in and the number of datasets uploaded to ONEShare, however this does not give us metrics that allow comparison of add-in versus web application. More broadly, we are not able to extensively evaluate the use of DataUp because of our limited ability to access user information. Based on the number of submissions to ONEShare, uptake of the tool by researchers has been minimal but steady. There is no related work section. The only related work is RightField (reference17) but what RightField does and how it relates to DataUp is not mentioned. Similar tools to DataUp such as ISAtool Suite and Ontomaton are not mentioned.We have added a paragraph to the introduction on existing work in this area.The survey appears USA-centric. EZID is only available in the USA. DataUp currently works with DataONEShare. Can and how DataUp be adapted for use in other repositories? Can it reuse infrastructure that is not US based? The code is openly available for anyone to use, and the CDL encourages other organizations to take the code and deploy their own instances of DataUp. The identifier provided for a dataset is via the repository; ONEShare is a US repository with connections to the EZID identifier service via the CDL. If other repositories deployed instances of DataUp, the identifier schema would be specific to their existing system." } ] }, { "id": "5385", "date": "09 Jul 2014", "name": "Louise Corti", "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 describes, in some detail, the development, implementation and piloting of a data documentation tool for tabular data.Overall the paper is very clear and provides an easy-to-read narrative of why and how the tool was developed, and how it can be used, as well as tips on populating and handling spreadsheet data, so the data has a longer-life.I am impressed with the simplicity of this tool, which attempts to solve issues in data description for a single type of data. This is much better than the 'workbench' approaches that try to do too much and end up failing.The issue of errors in data conversion between formats is critical and is a known issue in the data archival world. This tool addresses some of these common issues that arise in both spreadsheet data description and conversion.The paper presents some very useful information gathered from surveys and pilot work, but this is rather US-centric. I don't imagine the use cases of these type of scientists' behaviour in other countries are that different, but I would expect some pointers to this wider context.The referenced literature is good and covers many of the key sources I would refer to. However my own organisation in the UK has been advising on data documentation, including use of Excel and conversion issues, for some years, so it would be good to cite some examples of other efforts to address these issues on the non-ecology field and offer examples of non US resources that provide extensive data management advice (http://ukdataservice.ac.uk/manage-data.aspx).On page 6 the checklist of issues is very clear and useful and great to alert researchers to these issues upfront.In terms of platforms for the tool, I think a Mac version will be important. In my experience, many data creators prefer to have the convenience of local tools to document data, rather than relying on web-based tools, that can suffer from browser issues and loss of data through poor connection.I do believe that data preparation tools are best built into researchers' existing data handling software, as this brings the activities a step closer to data analysis and away from the burden of completing data deposit forms.I love the idea that the source code has been made available and that, on the whole, the project has been carried out in the spirit of openness, despite using a Microsoft base for the tool.I am also terribly impressed with the work done to convince Microsoft of the importance of this tool, and to secure codevelopment to enable it to be a plug-in. On this front, I have had some negative experience in lobbying software suppliers of qualitative analysis packages to implement a data exchange standard to enable export of within-system documentation; conversion between different market leaders' softwares is currently difficult, if not impossible. They should possibly take a leaf out of Microsoft's book and also listen to what data archivists/publishers are saying! The tool looks like it has had some user testing and feedback. Overall I believe this tool could have much wider value than the purposes for which the team have developed it. By simply replacing the metadata standard in use it could easily be applied to other disciplines, e.g. social science data. I would be very keen to pilot it and offer feedback on our own tabular data collection in the social sciences domain. The social sciences use the Data Documentation Initiative (DDI) which has fields that map pretty close to the schema used in the tool and discussed in this paper. I would advocate engagement with more data centres, possibly through forums like the Research Data Alliance.", "responses": [ { "c_id": "947", "date": "18 Aug 2014", "name": "Carly Strasser", "role": "Author Response", "response": "Reviewer text is italicized; our responses are below the reviewer text.The paper presents some very useful information gathered from surveys and pilot work, but this is rather US-centric. I don't imagine the use cases of these type of scientists' behaviour in other countries are that different, but I would expect some pointers to this wider context.In retrospect, it would have been valuable to collect use cases and feedback from non-US researchers. We were limited somewhat by our 12 month timeline and focused primarily on the communities that we could easily interview given our spatial and temporal constraints. Based on conversations with our international colleagues, it is our experience that researcher behaviors, concerns, and use cases are, across the board, fairly similar to what we found.However my own organisation in the UK has been advising on data documentation, including use of Excel and conversion issues, for some years, so it would be good to cite some examples of other efforts to address these issues on the non-ecology field and offer examples of non US resources that provide extensive data management advice (http://ukdataservice.ac.uk/manage-data.aspx).We recognize that many groups work in the area of best practices for data management. We did not, however, choose to focus on this aspect of community involvement since it has been largely covered by others. That said, we have added a paragraph referencing other projects that have similar goals to DataUp, one of which is based in the UK.In terms of platforms for the tool, I think a Mac version will be important. In my experience, many data creators prefer to have the convenience of local tools to document data, rather than relying on web-based tools, that can suffer from browser issues and loss of data through poor connection.We concur that a Mac version of the tool would be quite valuable. However, given our limited budget and abbreviated schedule, it was not possible for the first iteration of DataUp. Also, the fundamentally different architectures of the Mac and Windows versions of Excel meant that a Mac plug-in would have had to been created from scratch with no opportunity for code re-use from the Windows version. We hope that interested members of the open source will take on this task in the near future. We would be happy to support any such effort to the fullest extent possible.Overall I believe this tool could have much wider value than the purposes for which the team have developed it. By simply replacing the metadata standard in use it could easily be applied to other disciplines, e.g. social science data. I would be very keen to pilot it and offer feedback on our own tabular data collection in the social sciences domain. The social sciences use the Data Documentation Initiative (DDI) which has fields that map pretty close to the schema used in the tool and discussed in this paper. It’s true that the value of the tool goes well beyond just our target audience for requirements development and feedback. Social scientists in particular have expressed interest in the DataUp tool as a potentially valuable addition to their toolkit. One of the advantages of the DataUp/Dash merger is that the Dash platform has a more open architectural design that will greatly facilitate the process of supporting alternative metadata schemas.I would advocate engagement with more data centres, possibly through forums like the Research Data Alliance. We also agree that engaging more data centres would be ideal. The RDA was not yet formed when this work was conducted, but moving forward we hope to take advantage of such coalitions to get more uptake of a tool like DataUp." } ] } ]	1	https://f1000research.com/articles/3-6
https://f1000research.com/articles/3-2/v1	06 Jan 14	{ "type": "Case Report", "title": "Testicular artery pseudoaneurysm: a case report", "authors": [ "William P Parker", "Ajay K Nangia", "Ajay K Nangia" ], "abstract": "This is a case of an unusual cause of a testicular mass and the clinical features associated with its presentation and management. The patient presented with testicular pain and was found to have a testicular mass on ultrasound with a central 1cm anechoic region with arterial wave-form concerning for a pseudoaneurysm. The patient underwent orchiectomy with resolution of his symptoms. This case highlights the presentation of testicular artery pseudoaneurysm and outcome following orchiectomy.", "keywords": [ "A 34 year-old Caucasian man presented to the Emergency Department at the University of Kansas Medical Center (Kansas City", "Kansas) with a 1 week history of right-sided orchalgia in December of 2012. The day prior to presentation the patient was admitted at an outside institution for presumed orchitis", "where he was treated with intravenous levofloxacin 500mg. During that hospitalization a scrotal ultrasound was obtained which revealed hypervascularity of the right testicle but no masses or lesions within the testicle. He was discharged the next day", "but due to persistent symptoms presented to our institution. He denied any constitutional symptoms and was voiding without difficulty or irritative symptoms. He additionally denied any prior genitourinary trauma or infections including any history of orchitis", "prostatitis", "urinary tract infection", "or sexually transmitted infection. The patient was monogamous without any high-risk behavior. The patient had no past medical history and was on no prior medications. His examination revealed a swollen and indurated right testicle without involvement of the paratesticular structures. There was no discrete testicular mass palpable. Scrotal ultrasound revealed a region of hypoechogenicity measuring 3.3cm × 2cm felt to represent an intratesticular hematoma (see Figure 1). Within this there was a 1cm central focus that demonstrated an arterial wave form with alternating reversal of flow", "suggestive of a pseudoaneurysm (see Figure 2). Our differential diagnosis included abscess", "testicular artery aneurysm/pseudoaneurysm", "or testicular neoplasm. We counseled the patient on the ultrasonographic findings and our presumed diagnosis of testicular artery pseudoaneurysm. The patient was in a considerable amount of pain and was additionally concerned about the possibility of a testicular neoplasm and elected to undergo radical orchiectomy. Tumor markers were obtained prior to orchiectomy and were within normal limits. He tolerated the procedure without any adverse event and was discharged to home with resolution of his pain on the first post-operative morning. Pathologic evaluation confirmed the presence of significant intraparenchymal hemorrhage within a background of chronic orchitis (see Figure 3). At the time of post-operative follow-up – 2 weeks after his orchiectomy – the patient was in excellent condition with complete resolution of his pain. His surgical incision was well healed and he had no evidence of intra-scrotal pathology." ], "content": "Case history\n\nA 34 year-old Caucasian man presented to the Emergency Department at the University of Kansas Medical Center (Kansas City, Kansas) with a 1 week history of right-sided orchalgia in December of 2012. The day prior to presentation the patient was admitted at an outside institution for presumed orchitis, where he was treated with intravenous levofloxacin 500mg. During that hospitalization a scrotal ultrasound was obtained which revealed hypervascularity of the right testicle but no masses or lesions within the testicle. He was discharged the next day, but due to persistent symptoms presented to our institution. He denied any constitutional symptoms and was voiding without difficulty or irritative symptoms. He additionally denied any prior genitourinary trauma or infections including any history of orchitis, prostatitis, urinary tract infection, or sexually transmitted infection. The patient was monogamous without any high-risk behavior. The patient had no past medical history and was on no prior medications. His examination revealed a swollen and indurated right testicle without involvement of the paratesticular structures. There was no discrete testicular mass palpable. Scrotal ultrasound revealed a region of hypoechogenicity measuring 3.3cm × 2cm felt to represent an intratesticular hematoma (see Figure 1). Within this there was a 1cm central focus that demonstrated an arterial wave form with alternating reversal of flow, suggestive of a pseudoaneurysm (see Figure 2). Our differential diagnosis included abscess, testicular artery aneurysm/pseudoaneurysm, or testicular neoplasm. We counseled the patient on the ultrasonographic findings and our presumed diagnosis of testicular artery pseudoaneurysm. The patient was in a considerable amount of pain and was additionally concerned about the possibility of a testicular neoplasm and elected to undergo radical orchiectomy. Tumor markers were obtained prior to orchiectomy and were within normal limits. He tolerated the procedure without any adverse event and was discharged to home with resolution of his pain on the first post-operative morning. Pathologic evaluation confirmed the presence of significant intraparenchymal hemorrhage within a background of chronic orchitis (see Figure 3). At the time of post-operative follow-up – 2 weeks after his orchiectomy – the patient was in excellent condition with complete resolution of his pain. His surgical incision was well healed and he had no evidence of intra-scrotal pathology.\n\n\nDiscussion\n\nIntratesticular hemorrhage is frequently associated with trauma. In the absence of trauma, testicular artery aneurysm and pseudoaneurysm have been described as an infrequent source of hematoma formation. To our knowledge there have been two cases of testicular artery aneurysm and two of pseudoaneurysm reported in the literature, with etiologies of trauma1–3 and infection4.\n\nAs in previous cases of testicular artery aneurysm and pseudoaneurysm, the diagnosis was established on ultrasonography. Furthermore this case describes a second scenario in which sonographically diagnosed orchitis has progressed to this clinicopathologic entity4. Our management included radical orchiectomy - to rule out possible malignancy - and the patient recovered with complete resolution of his pain. This case supports orchitis as a risk factor for pseudoaneurysm formation, the use of ultrasound for the diagnosis, and the use of orchiectomy as a potential treatment in patients with unremitting pain.\n\nCertainly the management of this case is limited by the radical treatment offered – namely orchiectomy. We did consider and offer the patient a more conservative trial of observation for possible resolution. However in the absence of overwhelming evidence to rule out malignancy we felt that radical orchiectomy would not only establish our working diagnosis, but also rule out the possibility of testicular neoplasia. Regardless, at the time of final follow-up the patient was satisfied with his course of care and the outcome.\n\n\nConsent\n\nThe patient was unable to be reached for consent and no next-of-kin information was available to contact the patient. The write-up does not contain sufficient information to identify the patient as this is a case based mainly on radiology and pathology findings. We have made numerous attempts to contact the patient and it appears that his contact information as provided at the time of treatment is no longer valid.", "appendix": "Author contributions\n\n\n\nWilliam Parker prepared the manuscript; Ajay Nangia edited the manuscript and participated in the clinical care of the patient.\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\nDee KE, Deck AJ, Waitches GM: Intratesticular pseudoaneurysm after blunt trauma. AJR. 2000; 174(4): 1136. PubMed Abstract \| Publisher Full Text\n\nZicherman JM, Mistry KD, Sarokhan CT, et al.: CT angiography, sonography, and MRI of aneurysm of the testicular artery. AJR. 2004; 182(4): 1088–1089. PubMed Abstract \| Publisher Full Text\n\nReddy YP, Murphy JK, Sheridan WG: Spontaneous aneurysm of the testicular artery. Br J Urol. 1998; 82(4): 599–600. PubMed Abstract \| Publisher Full Text\n\nMujoomdar A, Maheshwari S, Zand KR, et al.: Sonographic diagnosis of a ruptured intratesticular pseudoaneurysm secondary to orchitis. AJR. 2007; 189(1): W20–22. PubMed Abstract \| Publisher Full Text" }	[ { "id": "2993", "date": "08 Jan 2014", "name": "Nelson Bennett", "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 very nice presentation and description of a rare issue. The manuscript underscores that a complete scrotal/testicular ultrasound be carefully scrutinized. To further enhance the impact of the study, the authors may want to consider including the values of the testicular tumor markers.Good Show.", "responses": [] }, { "id": "2990", "date": "24 Jan 2014", "name": "Paul Turek", "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 case report is a classic “suspicious-looking-lesion-in-the-testicle” that is treated by removal and diagnosed definitively only after the fact, based on histology. The care provided was standard. However, besides learning about the existence of the lesion, there is no information in this case report that might help others PREVENT the extreme measure of radical orchiectomy for a benign lesion. What other diagnostic tests could have been done? Why wasn't partial orchiectomy considered in the treatment algorithm? The learning potential of this case is limited.", "responses": [] }, { "id": "5013", "date": "16 Jun 2014", "name": "Jay I. Sandlow", "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 fairly rare case of testicular artery pseudoaneurysm, presumably secondary to orchitis. The case history is complete and the figures are well-presented. Furthermore, based on the patient’s history, age, and ultrasound findings, their management was not unreasonable. However, I would like them to describe in more detail the thought process that led to a radical orchiectomy, particularly since pseudoaneurysm was in their differential. Because the patient was of reproductive age, I wonder if it would have been possible to preserve some functioning testicular tissue. Based on the gross appearance of the specimen, it appears that at least a third of the testis was unaffected. How accurate is scrotal ultrasound in identifying a testicular artery pseudoaneurysm? Finally, looking back, is there something that the authors could have done to prevent a radical orchiectomy? This might be helpful for those of us facing this situation in the future. I think with some revisions of the Discussion, this case report should be acceptable.", "responses": [] } ]	1	https://f1000research.com/articles/3-2
https://f1000research.com/articles/3-219/v1	12 Sep 14	{ "type": "Review", "title": "Aging and energetics’ ‘Top 40’ future research opportunities 2010-2013", "authors": [ "David B. Allison", "Lisa H. Antoine", "Scott W. Ballinger", "Marcas M. Bamman", "Peggy Biga", "Victor M. Darley-Usmar", "Gordon Fisher", "Julia M. Gohlke", "Ganesh V. Halade", "John L. Hartman", "Gary R. Hunter", "Joseph L. Messina", "Tim R. Nagy", "Eric P. Plaisance", "Mickie L. Powell", "Kevin A. Roth", "Michael W. Sandel", "Tonia S. Schwartz", "Daniel L. Smith", "J. David Sweatt", "Trygve O. Tollefsbol", "Stephen A. Watts", "Yongbin Yang", "Jianhua Zhang", "Steven N. Austad", "David B. Allison", "Lisa H. Antoine", "Scott W. Ballinger", "Marcas M. Bamman", "Peggy Biga", "Victor M. Darley-Usmar", "Gordon Fisher", "Julia M. Gohlke", "Ganesh V. Halade", "John L. Hartman", "Gary R. Hunter", "Joseph L. Messina", "Tim R. Nagy", "Eric P. Plaisance", "Mickie L. Powell", "Kevin A. Roth", "Michael W. Sandel", "Tonia S. Schwartz", "Daniel L. Smith", "J. David Sweatt", "Trygve O. Tollefsbol", "Stephen A. Watts", "Yongbin Yang", "Jianhua Zhang" ], "abstract": "Background: As part of a coordinated effort to expand our research activity at the interface of Aging and Energetics a team of investigators at The University of Alabama at Birmingham systematically assayed and catalogued the top research priorities identified in leading publications in that domain, believing the result would be useful to the scientific community at large.Objective: To identify research priorities and opportunities in the domain of aging and energetics as advocated in the 40 most cited papers related to aging and energetics in the last 4 years.Design: The investigators conducted a search for papers on aging and energetics in Scopus, ranked the resulting papers by number of times they were cited, and selected the ten most-cited papers in each of the four years that include 2010 to 2013, inclusive.Results:\n\nTen research categories were identified from the 40 papers. These included: (1) Calorie restriction (CR) longevity response, (2) role of mTOR (mechanistic target of Rapamycin) and related factors in lifespan extension, (3) nutrient effects beyond energy (especially resveratrol, omega-3 fatty acids, and selected amino acids), 4) autophagy and increased longevity and health, (5) aging-associated predictors of chronic disease, (6) use and effects of mesenchymal stem cells (MSCs), (7) telomeres relative to aging and energetics, (8) accretion and effects of body fat, (9) the aging heart, and (10) mitochondria, reactive oxygen species, and cellular energetics.Conclusion: The field is rich with exciting opportunities to build upon our existing knowledge about the relations among aspects of aging and aspects of energetics and to better understand the mechanisms which connect them.", "keywords": [ "Energetics can be defined as the study of the causes", "mechanisms", "and consequences of the acquisition", "storage", "and utilization of metabolizable energy by biological organisms. The United States – indeed the world – is currently undergoing a crisis of excess energy storage", "sometimes called the obesity epidemic. At the same time", "the United States population is aging at an unprecedented rate with the over 65 population expected to double by 2050", "the over 85 population expected to triple", "and the over 100 population expected to more than sextuple (U.S. Administration on Aging", "http://www.aoa.gov/Aging_Statistics/future_growth/future_growth.aspx#age). A consistent finding from ecology", "basic laboratory science", "and epidemiologic research is that aspects of energetics", "including the perceived and actual availability of food", "the ingestion of food", "the composition of the food consumed", "the amount of body energy accreted and expended", "affect disease and disability", "senescence", "mortality rate", "and longevity. A team of investigators at The University of Alabama at Birmingham (UAB) is currently aiming to advance innovative research at the interface of aging and energetics (hereafter A&E)", "and in preparation took stock of what ‘the field’ was describing as the vital research needs and opportunities in this domain. Realizing that the results of this assessment may be useful to others working in A&E", "we publish them here." ], "content": "Introduction\n\nEnergetics can be defined as the study of the causes, mechanisms, and consequences of the acquisition, storage, and utilization of metabolizable energy by biological organisms. The United States – indeed the world – is currently undergoing a crisis of excess energy storage, sometimes called the obesity epidemic. At the same time, the United States population is aging at an unprecedented rate with the over 65 population expected to double by 2050, the over 85 population expected to triple, and the over 100 population expected to more than sextuple (U.S. Administration on Aging; http://www.aoa.gov/Aging_Statistics/future_growth/future_growth.aspx#age). A consistent finding from ecology, basic laboratory science, and epidemiologic research is that aspects of energetics, including the perceived and actual availability of food, the ingestion of food, the composition of the food consumed, the amount of body energy accreted and expended, affect disease and disability, senescence, mortality rate, and longevity. A team of investigators at The University of Alabama at Birmingham (UAB) is currently aiming to advance innovative research at the interface of aging and energetics (hereafter A&E), and in preparation took stock of what ‘the field’ was describing as the vital research needs and opportunities in this domain. Realizing that the results of this assessment may be useful to others working in A&E, we publish them here.\n\n\nMethods/Approach\n\nTo identify the ‘Top 40’ Articles in A&E, we conducted the following search in Scopus: ((TITLE((aging OR ageing OR lifespan OR longevity OR senescence)) AND PUBYEAR > 2009 AND PUBYEAR < 2014) AND (TITLE((calori* OR diet* OR energetic* OR nutri* OR food OR fat OR adipo* OR \"body composition\")) AND PUBYEAR > 2009 AND PUBYEAR < 2014)) AND (LIMIT-TO(LANGUAGE, \"English\")).\n\nWe then ranked the papers by number of times they were cited and selected the ten most cited papers in each of the four years that include 2010, 2011, 2012, and 2013. This ‘normalized’ for amount of time to be cited and resulted in a total of 40 papers (Appendix A).\n\nTen of the authors then each reviewed 4 of the 40 top A&E publications and identified areas relative to A&E that were recommended by the original papers’ authors as needs for future research. For example, we extracted statements from the publications that included phrases like “More studies are needed on…”, “Future work should…”, “Further studies are needed to …”, “…not yet been established …”, and so on. Thereafter, two authors (DBA and LHA) grouped these statements into thematic categories. This involved some subjectivity and loss of some recommendations which did not easily fit in any category, but seemed to capture the vast majority of the recommendations. The ten thematic categories were labeled: 1)CR longevity response, 2) role of mTOR and related factors in lifespan extension, 3) nutrient effects beyond energy (especially resveratrol, omega-3 fatty acids, and selected amino acids, 4) autophagy and increased longevity and health, 5) aging-associated predictors of chronic disease, 6) use and effects of MSCs, 7) telomeres relative to aging and energetics, 8) accretion and effects of body fat, 9) the aging heart, and 10) mitochondria, reactive oxygen species, and cellular energetics. Specific suggestions often appeared in more than one category.\n\nWe now briefly summarize the recommendations in each category.\n\n\nCR longevity response\n\nIt was in rats, in 1917, that caloric restriction (CR) was first documented to extend life1. In the decades following, researchers studying model organisms as diverse as yeast, Drosophila, rodents, and primates have accumulated experimental evidence demonstrating that CR extends lifespan, slows the aging process, and/or improves healthy aging biomarkers2–4. Concurrently, repeated studies within select model species (e.g., rats) further supported the robustness of this response4. And thus, with this mounting evidence, the anti-aging response to a protocol of 20 to 40% CR was largely considered to be a robust response, evolutionarily conserved across animals.\n\nIntrigued by this perceived adaptive response to an obvious physiological stress, researchers began traversing the animal lineage, beyond the traditional laboratory model organisms, to validate the robustness of this response. Instead they found extensive heterogeneity across species in their ability to extend median or maximum lifespan in response to CR4. Similarly, testing additional genetic strains of lab animals uncovered extensive variation in the response to CR within inbred, outbred, and wild-derived strains of nematodes and mice (currently over 80 strains tested)4–6. For example, while CR has a relatively robust response in rats and C. elegans, by causing nearly all strains tested to increase median or maximum lifespan (~30% increase in rats); in mice this is true for only a subset of strains4. Rather, the effect of CR across mouse strains (particularly the ILSXISS) ranged from decreasing or having no effect to increasing lifespan, with an average increase of only ~15%4. Furthermore, this heterogeneity may be underplayed due to bias in reporting the negative effects of CR on lifespan4. Thus, despite the anti-aging response to CR being evolutionarily conserved, there is now considerable evidence for heterogeneity in either the presence or absence of the lifespan extension response, and the degree of the effect on lifespan.\n\nDespite this variation, researchers are actively searching for conserved molecular mechanisms underlying this response, typically using the short-lived model organisms. Many molecular mechanisms are implicated in this response, but few have been or are found to be consistent across species. A generalized theme is that reprogramming of metabolism in the face of CR is key to achieving lifespan extension2.\n\nThe current challenge is to understand the sources of this heterogeneity and to determine if the underlying molecular mechanisms identified in the short-lived model organisms can translate to humans2. The heterogeneity of the CR response across mice strains depicts a role for genetic variation5. Some of the variation among strains is linked to the ability to regulate fat metabolism, thereby highlighting the link between body composition and aging5,7. The genetic variation also points to the possibility for natural selection to act on this response, which may explain the heterogeneity across populations and species of non-model organisms. The variable results in repeated studies within a genetic strain points to exceptional sensitivity of the response to other environmental variables. Indeed mammals under CR are often more susceptible to pathogens and other environmental stresses (e.g., cold stress) which may vary among laboratories and studies4. Finally, heterogeneity among strains likely results from differences in gene by environment interactions, such that the optimal CR protocol (e.g., level of CR, dietary composition, developmental timing of implementation, and consistency of the regime) for producing anti-aging effects may be unique for each genetic background or population4,7.\n\nOne important question to consider is what one might call ‘forced inactivity’ and consequent low energy expenditure and any healthspan/lifespan consequences in caged laboratory animals. While there is substantial heterogeneity in volume of voluntary exercise across strains8, the main point is that lab mice will voluntarily run some amount (and up to 11 km/day) if given a wheel, and may reap healthspan benefits from it. The field as a whole may create animals (particularly rodents) predisposed to disease and early death because of forced inactivity (see, e.g., 9–11).\n\nOverall, CR is the most discussed intervention for extension of life--span. While largely, but not perfectly, evolutionarily conserved across animals, there is extensive heterogeneity in this response within and across species. Understanding the heterogeneity in the response to CR at both the organismal and molecular level will be essential for predicting the plausible effects of elements of a CR regime or its pharmaceutical mimics to heterogeneous human populations.\n\nTraditional thinking about CR, developed primarily with laboratory rodents, held that a significant restriction of dietary calories relative to ad lib feeding universally increased longevity and that energy intake alone, rather than the amount of any particular dietary nutrient, modulated lifespan12,13. Recent research on an extended range of model organisms makes both of those general conclusions no longer tenable. Not only does the same CR regime that extends life in some genotypes of mice, flies, and yeast have no effect on longevity in other genotypes14–16, it even shortens life in some genotypes indicating less than perfect evolutionary conservation17,18. Also, new evidence clearly indicates that restriction of some nutrients, in particular methionine or protein generally, is sufficient for significant life extension across a range of organisms19–22. Consequently, in some cases, the term “CR” as a descriptor of a longevity-enhancing, reduced nutrient regime should probably be replaced with the more generic and previously widely-used term “dietary restriction (DR)”. The life-extending impact of nutrient reduction, where it is observed, can be reasonably referred to as the “DR effect”.\n\nWhat can these new studies, employing a broader array of species, genotypes, and nutrient regimes than previously, suggest about mechanisms of life extension by dietary restriction? First, the lack of a life-extending effect, or even a life-shortening effect, of DR in some genotypes can be informative. For instance, in a series of recombinant inbred mouse strains, the ability of a strain to live longer under 40% CR was significantly heritable and related to metabolic efficiency5. Mice that lost the least weight and preserved the most body fat under reduced calories were most likely to live longer7. These results together suggest that allelic variation in energetic efficiency genes may underlie the DR effect. Similarly, variation in replicative longevity in a series of 166 yeast single-gene deletion strains implicated vacuolar pH, superoxide dismutase activity, and mitochondrial proteome homeostasis (= proteostasis) as key determinants of the response to DR. Follow-up studies in C. elegans, which has been evolutionarily separated from yeast for ~1 billion years, also implicate mitochondrial proteostasis in the DR effect18. Together these findings suggest a key role for energetic efficiency as a central process for understanding how nutrient intake and processing affect health and longevity. Second, despite the promise of a unified, conserved mechanism for the DR effect suggested above, other evidence indicates more mechanistic variety. For instance, the growth hormone (GH)/IGF-1 axis is strongly implicated in the DR effect in mice such that: (1) reduced GH and IGF-1 signaling have been genetically associated with longer life, (2) DR in mice reduces signaling through these pathways, and (3) DR fails to increase longevity substantially in GH receptor knockout mice23. By contrast, in C. elegans insulin/IGF signaling is not required for the DR effect – at least under most DR regimes24. However, DR is implemented by at least 12 methods in worms, differing in the culture medium, food source, and the age at which restriction is initiated. Surprisingly, the genetic pathways implicated in the DR effect vary depending upon method, suggesting considerable mechanistic complexity for such a seemingly straightforward environmental manipulation.\n\nWhat does this avalanche of new information on DR across species suggest for future research directions according to our most-cited recent papers? First, and with most consensuses, determining the impact on health and longevity of the relative intake of different nutrients as well as the timing and number of calories consumed needs refinement and re-evaluation across a range of species and genotypes2–4,17,25–30. However true, this suggestion lacks focus and more helpful perhaps is the linking by several papers of the DR effect either directly or indirectly to metabolic efficiency5,18,26 and the maintenance of proteostasis20,25,31,32. With respect to this latter process, the most commonly adduced mediator of proteostasis was modulation of mTOR activity2,25,26,32,33. Herein lies the most focused suggestion for future research to arise from our most-cited papers.\n\n\nRole of mTOR and related factors in lifespan extension\n\nThe mTOR kinase integrates signals from nutrients, energy status, growth factors, and a variety of stressors and affects, among other things, the rate of protein synthesis and degradation via autophagy by a host of downstream effectors34. Thus TOR inhibition is a major player in the management of cellular energy and the maintenance of the proteome with aging. A number of our top-cited papers emphasize that the link between health, aging, nutrition, and energetics would be dramatically advanced by increased knowledge of how various nutrient and caloric manipulations affect and integrate both upstream mediators and downstream effectors of TOR activity, particularly on a tissue-specific basis.\n\nRapamycin, a macrolide with immunomodulatory properties, and the product of the bacterium Streptomyces hygroscopius, was found to inhibit an evolutionarily conserved protein belonging to the phosphatidyl inositol kinase-related serine/threonine kinase family, referred to as mTOR, the mechanistic target of rapamycin2,25,26,35. Modulating the mTOR signaling network affects mRNA translation, transcription, autophagy, ribosomal biogenesis, metabolism and cell survival, proliferation, cell size and growth, endoplasmic reticulum stress signaling, and other stress responses26. In mammals, CR, or reduced dietary protein or essential amino acid intake, can extend longevity, improve metabolic fitness, and increase stress resistance, at least in part via inhibition of mTOR activity25. Nutrients (glucose, amino acids, especially leucine, and fatty acids) directly activate the mTOR pathway and also increase insulin levels, which can additionally activate mTOR35. mTOR acts as a major signaling hub, integrating multiple inputs and mediating the switch between growth and somatic maintenance to extend lifespan26. Mammalian cells have a single mTOR kinase which exists in two structurally and functionally distinct multiprotein complexes: mTORC1 and mTORC2. mTORC1 is acutely inhibited by rapamycin, although chronic rapamycin administration inhibits mTORC2 as well36. Amino acid deprivation is also a powerful inhibitor of mTORC1 activity even the presence of acute growth factor stimulation25,26. Strong inhibition of mTORC1 early in life drastically slows or even stops development, but administration of rapamycin or other inhibitors of mTORC1 late in life causes lifespan extension, boosts immune function and rejuvenates hematopoietic stem cells25,26.\n\nThe specific upstream mechanisms directly responsible for amino acid sensing which are permissive of mTOR activation remains unresolved, but localization of the mTORC1 complex has emerged as an important aspect of amino acid mediated control25 (see Figure 1 and Figure 2). The mTORC1 pathway controls protein synthesis most directly by phosphorylating and inhibiting a repressor of cap-dependent mRNA translation, 4E-BP1. mTORC1 also regulates translation indirectly through activation of S6 kinase (S6K) whose substrate is the ribosomal protein S6, a component of the 40S ribosome important for translation26.\n\na \| Growth factors such as insulin stimulate PI3K to generate phosphatidylinositol-3,4,5-triphosphate (PtdIns(3,4,5)P3), which promotes the phosphorylation (P) of AKT at Thr308 by phosphoinositide-dependent kinase 1 (PDK1). AKT phosphorylates tuberous sclerosis complex 2 (TSC2) on multiple sites to inhibit its GTPase-activating protein (GAP) activity for the small GTPase RAS homologue enriched in brain (RHEB). GTP-loaded RHEB then activates mammalian TOR complex 1 (mTORC1). Growth factors also stimulate mTORC2 by promoting its association with ribosomes in a PI3K-dependent manner. b \| Amino acids stimulate mTORC1 by promoting the conversion of RAS-related GTP-binding protein (RAG) heterodimers to the active conformation, in which RAGA or RAGB is loaded with GTP and RAGC or RAGD is loaded with GDP. c \| In response to low energy (high AMP/ATP ratio), AMP-activated protein kinase (AMPK) phosphorylates regulatory-associated protein of mTOR (RAPTOR) at Ser792 and TSC2 at Ser1387, leading to the inhibition of mTORC1. d \| During the inhibition of the Hippo pathway component large tumour suppressor homologue (LATS) kinase, hypophosphorylated Yes-associated protein (YAP) translocates to the nucleus and promotes the expression of the microRNA miR-29. miR-29 targets PTEN mRNA and inhibits PTEN translation, which leads to increased levels of PtdIns(3,4,5)P3 and the activation of both mTORC1 and mTORC2. Dashed arrows represent translocation of the molecule. e \| Glycogen synthase kinase 3β (GSK3β) activates the TSC complex by phosphorylating TSC2 at Ser1379 and Ser1383. Phosphorylation of these two residues requires priming by AMPK-dependent phosphorylation of Ser1387. WNT signalling inhibits GSK3β and the TSC complex, and thus activates mTORC1. mTORC2 is activated by WNT in a manner dependent on the small GTPase RAC1. Proteins shown in green promote mTOR activity or their activity is promoted by mTOR. Proteins shown in red inhibit mTOR activity or their activity is inhibited by mTOR. Phosphorylation depicted in yellow is an activation signal and phosphorylation depicted in orange is an inhibitory signal. 4E-BP, eIF4E-binding protein; APC, adenomatous polyposis coli; GPCR, G protein-coupled receptor; GRB10, growth factor receptor-bound protein 10; IGF, insulin-like growth factor 1;IRS1, insulin receptor substrate 1; LRP, low-density lipoprotein receptor-related protein; mLST8, mammalian lethal with SEC thirteen 8; RICTOR, rapamycin-insensitive companion of mTOR; S6K, ribosomal S6 kinase; SIN1, SAPK-interacting 1; TBC1D7, TBC1 (TRE2–BUB2–CDC16) domain family member 7; ULK1, UNC-51-like kinase 1.\n\nThis figure has been reproduced with kind permission from Shimobayashi M, Hall M. Making new contacts: the mTOR network in metabolism and signalling crosstalk. Nature Reviews Molecular Cell Biology 2014;15:155–162.\n\nMammalian TOR complex 1 (mTORC1) promotes anabolic processes, such as the biosynthesis of proteins, nucleotides and lipids, and inhibits catabolic processes such as autophagy. Aa \| mTORC1 phosphorylates (P) the hydrophobic motif (Thr389) in ribosomal S6 kinase (S6K), thereby activating it to subsequently phosphorylate ribosomal protein S6 at the sites indicated to promote ribosome biogenesis. mTORC1 also phosphorylates eIF4E-binding protein (4E-BP) at multiple sites to inhibit it. Inhibition of 4E-BP stimulates translation initiation, especially of 5′ oligopyrimidine tract (termed a 5′ TOP) and pyrimidine-rich translational element (PRTE) containing mRNAs. Ab \| mTORC1 stimulates nucleotide and lipid synthesis. mTORC1 promotes the gene expression of key enzymes in the pentose phosphate pathway (PPP), at least in part by activating sterol regulatory element-binding proteins (SREBPs). mTORC1 also stimulates CAD (Gln-dependent carbamoyl-phosphate synthase, Asp carbamoyltransferase, dihydroorotase) by S6K-mediated phosphorylation at Ser1859, which leads to CAD activation and the stimulation of de novo pyrimidine synthesis. Furthermore, mTORC1 promotes lipogenic gene expression by activating S6K or by inhibiting the nuclear translocation of LIPIN1, both of which activate the transcription factor SREBP. Ac \| mTORC1 inhibits autophagy by phosphorylating UNC-51-like kinase 1 (ULK1) at Ser758 and ATG14 at multiple sites. During mTORC1 inhibition, AMPK phosphorylates ULK1 at Ser317, and thereby activates ULK1, which phosphorylates Beclin 1 in the vacuolar protein sorting 34 (VPS34)–Beclin 1–ATG14 complex to initiate autophagy. mTORC1 also inhibits autophagy indirectly by blocking lysosome biogenesis, by phosphorylating and inhibiting the nuclear translocation of transcription factor EB (TFEB). B \| mTORC2 co-translationally phosphorylates AKT at Thr450 to prevent its ubiquitylation and degradation. mTORC2 also post-translationally phosphorylates ATK at Ser473 to promote lipogenic gene expression by activation of SREBP1c. Moreover, mTORC2 co-translationally phosphorylates IGF2 mRNA-binding protein 1 (IMP1) at Ser181, which stimulates insulin-like growth factor 2 (IGF2) production. The activity of proteins shown in green is promoted by mTOR. The activity of proteins shown in red is inhibited by mTOR. Phosphorylation depicted in yellow is an activation signal and phosphorylation depicted in red is an inhibitory signal. Dashed arrows represent translocation of the protein. mLST8, mammalian lethal with SEC thirteen 8; PDK, phosphoinositide-dependent kinase 1; RAPTOR, regulatory-associated protein of mTOR; RICTOR, rapamycin-insensitive companion of mTOR; SIN1, SAPK-interacting 1.\n\nThis figure has been reproduced with kind permission from Shimobayashi M, Hall M. Making new contacts: the mTOR network in metabolism and signalling crosstalk. Nature Reviews Molecular Cell Biology 2014;15:155–162.\n\nA drop in the cell’s energy content is reflected in the rise of the AMP/ATP ratio activating AMPK, which reduces the activity of mTORC1 by direct phosphorylation of TSC2 (Tuberous Sclerosis Complex 2), a negative regulator of mTORC1, and Raptor, an essential binding partner for the induction of translation by mTORC12,26. Biochemical evidence places AMPK upstream of mTORC1, but AMPK has also been shown to be downstream of S6K and can mediate S6K dependent effects on lifespan. Similarly, S6K is directly downstream of mTORC1, but via its inhibitory effect on IRS1 and 2, it is also upstream of mTORC1 via an insulin-signaling route34. The complex biochemical interactions among the various players in the mTOR network clearly require more refined mechanistic investigation26. Thus, mTORC1, S6K and AMPK may constitute a complex feedback loop sensitive to dietary restriction which can redirect growth, metabolism and lifespan. In addition to nutrients and growth factors, appropriate control of growth also requires integration of information on environmental stresses. Osmotic stress, hypoxia, ER stress, genotoxic stress, mechanical forces, and contraction (e.g., muscle movement) may all regulate mTORC1 activity26.\n\nKey questions that still need to be delineated are how multiple nutrients, growth factors and stress inputs are integrated into a single mTORC1 activity level; how the physiological responses downstream of mTORC1 are differentially regulated; how mTOR mediates changes in survival and lifespan in response to various environmental manipulations, especially nutrients; how the sex-dependent effects of mTORC1 signaling on longevity are regulated; with multiple mechanisms proposed; how mTORC1 activity couples nutrient availability to ribosome biogenesis; and which transcription factors involved in mediating stress responses are regulated by mTORC1 signaling to affect lifespan extension? Furthermore, the downstream effectors of mTORC1 are not sufficiently determined to understand which pathways are absolutely necessary for mTORC1’s role in lifespan extension. The role of mTORC1 is likely to be highly tissue specific, and the communication between cells, tissues, and organs to determine lifespan is a major knowledge gap. Lastly, little is known about the role of mTORC1 in primate aging and life extension by CR in those primate populations in which such extension is seen2,25,26,28,35.\n\n\nNutrient effects beyond energy (especially resveratrol, omega-3 fatty acids, and selected amino acids)\n\nIn addition to the well-recognized effect of CR on longevity and disease, there is growing evidence that the abundance of individual nutrients, macronutrient types, and nutritional supplements (i.e., non-caloric dietary compounds) specifically influence health and longevity. For example, restriction of the essential amino acid, methionine, can improve health and increase longevity in mice and rats, although importantly too much restriction can compromise health as well21,37,38. Methionine restriction provides benefits independent of caloric intake (i.e., with ad lib feeding). Although mechanistic differences are not fully understood, one possibility is that specific pathways sensitive to dietary nutrients can be targeted to alter lifespan. Examples include amino acid signaling with the mTOR pathway, energy status sensing through the adenosine monophosphate kinase (AMPK) pathway, oxidative stress through SKN1/Nrf2, protein quality control though enhanced autophagy and the general stress response. A variety of perturbations may activate ‘longevity’ pathways such as these to promote better health.\n\nThere is also evidence accumulating that plant-based, bioactive compounds (phytochemicals) have a variety of cellular effects and also work through multiple pathways. While a focused approach is helpful for establishing cause-effect for individual compounds, it is becoming clear that we must understand combinations of dietary effects and their interactions to fully describe the biological complexity underlying their contributions to homeostatic control and cellular regulation. For instance, one of the most well-known, and highly-researched CR mimetics (compounds which are proposed to mimic the benefits of CR without the necessity of energy intake reduction) is resveratrol, a polyphenol found in the skins of red grapes, muscadines, etc. Resveratrol is proposed to possess multiple agonist (sirtuins, AMPK) and antagonist (TOR/S6K, heat shock response) activities to pathways which have been genetically demonstrated to mediate longevity and health outcomes. Yet, the effects of resveratrol remain open to question. Resveratrol has not yielded consistent results on longevity outcomes, but it remains a highly researched phytochemical that may mediate disease risk or cellular/tissue health through other of these multiple interactions. Mice appear “largely unaffected by resveratrol”40.\n\nAn alternative approach to pathway specific targeting is exemplified in omega-3 fatty acid research. Polyunsaturated fatty acids (PUFAs), which contain double bonds in their carbon backbone, can be subdivided in various families, with the ‘omega-3’ (double bond at the n-3 position) type PUFA garnering the most health claims. This may be related to the change in the structure and biochemical properties of the lipid that provides important components of cellular membranes and protection from oxidative stress, particularly in the central nervous system where omega-3s are concentrated in nervous tissues. Although dietary deficiency has been associated with neurodegenerative disease risk and accelerated aging, supplementation with omega-3s (above the accepted minimum requirement) to slow aging and reduce disease requires further testing. In fact it is a common conundrum in the nutrition/metabolism research field as to what constitutes nutritional deficiency or excess, and whether compensatory nutritional alterations (supplementation or restriction of the ‘required’ amount, or simply removing an ‘excess’ or ‘deficiency’) affect healthspan or lifespan. Thus, by definition, causes and mediators of aging require not just the alleviation of a disease inducing state, but promotion of a health maintaining state that reduces disease risk, while increasing health span and longevity. To help address this question, a recent study investigated the effect of macronutrient balance on longevity in mice by testing 10 diet compositions at 3 energy densities, and found that mTOR signaling responded to the dietary protein content (increased mTOR signaling with more protein), raising the possibility that compounds like rapamycin which have significant longevity effects and alter amino acid signaling, may be modified by the background dietary protein and/or amino acid content. Alternatively, varied dietary compositions having different macronutrient proportions, individual nutrient restrictions, impaired nutrient metabolism, etc. could be modified to reduce signaling through pro-aging pathways, thereby increasing health and longevity. Whether this is the underlying mechanism(s) by which CR exerts its varied health benefits requires future study. However, manipulations of the diet by compound inclusion or individual nutrient restrictions, because they permit unrestricted intake, remains a type of intervention that more people will likely accept than CR as a change in their lifestyle.\n\n\nAutophagy and increased longevity and health\n\nAging and senescence are associated with accumulation of damaged proteins and organelles, mitochondrial dysfunction, and metabolic imbalance, especially in post-mitotic cells41–48. These pathologies further exacerbate cell injury, resulting in organ failure and decreased lifespan. While the oxidative stress theory of aging has been explored, except in limited examples using gene knockout or overexpression of antioxidant enzymes, there is a general lack of correlation between the levels of antioxidants and lifespan49–55. Since proteins and organelles are normally cleared by the autophagy-lysosomal activity, an emerging concept for regulation of aging and lifespan is that autophagy-lysosomal activity becomes deficient in aging32,56–67.\n\nIn general, it has been shown that autophagy is regulated at many levels, by nutrient availability, oxidative and reductive stress, hypoxia and pathogenic molecules68,69. When this activity is unable to meet the demand for clearance of damaged intracellular macromolecules and organelles, pathologies emerge and progress. For example, perturbation of autophagy is evident in neurodegenerative diseases including Alzheimer’s and Parkinson’s diseases70,71. In yeast, worms, flies, and mammals, insufficient autophagy contributes to accumulation of protein aggregates and dysfunctional mitochondria, thereby leading to a decrease in lifespan and the pathogenesis and progression of various age-dependent chronic diseases56,57,59. In addition, inhibition of autophagy leads to deterioration in mitochondrial function72,73; for example, ATG7-deficient skeletal muscles and pancreatic β cells accumulate dysmorphic mitochondria and exhibit defective oxidative phosphorylation73. In addition PINK1 knockout mice exhibit mitochondrial dysfunction in striatum, liver, brain tissue, and primary cortical neurons74,75.\n\nIn a complementary series of studies, a role for the beneficial effects of stimulating autophagy in healthy aging has emerged based on data from longevity studies in various models31,60–62,64,65,76,77. Rapamycin, which inhibits mTORC1 activity and activates autophagy, has been shown to extend lifespan in several model organisms76–78. CR and resveratrol have also been shown to extend lifespan through SIRT1-mediated activation of autophagy in worms31, and genetic manipulations of autophagy also extend lifespan in animal models57,60,62,64,65.\n\nWhile autophagy is an attractive therapeutic target for increasing life and health span, key questions remain regarding our understanding of the fundamental mechanisms of autophagic regulation, as well as the impact of autophagy activation on lifespan and health span. Investigations of these key questions are essential to our ability to fully exploit this pathway for improving healthy aging, longevity, and associated drug development.\n\nOne critical aspect is already clear – that the fundamental mechanisms of autophagy regulation are context-dependent. For example, autophagic activation in response to oxidative stress is differentially regulated from autophagy activated by starvation. Determining the relevant common and specific autophagy mechanisms, and identifying the signaling pathways that modulate the level, location and selectivity of autophagy, are crucial for a better understanding of autophagy in aging. Regarding our understanding of the impact of autophagy on aging, age-related disease and longevity, we must take into account that current animal studies are predominantly undertaken using relatively pathogen- and nemesis-free environments. How we translate these studies to humans is still an unanswered question. Epidemiologic studies can give important clues in this regard, although pinpointing a specific contribution of autophagy on life and health span in human longevity studies remains a technical challenge.\n\nAn interesting and topical aspect is the extent to which individual variation in the autophagy pathway impacts healthy aging. Understanding the genetic variants in the human population which modulate autophagy is a largely unexplored aspect of genomic studies, yet will be essential as we enter the era of personalized health care.\n\nPerhaps metabolomics, proteomics, or bioenergetics health indices need to be developed using imaging, accessible biopsies or body fluid to provide the much-needed paradigm shifts in our understanding of, and capacity for, monitoring human health conditions80–82. Additionally, autophagy activation may have differential impact on different organs in terms of their function and longevity. If improving health span is what we desire, then how to define health is not as simple as it appears. What one person defines as a healthy circumstance, another person may find unacceptable. Thus, we may have to be able to distinguish health and illnesses of different organs in any given person.\n\nThe emerging autophagy theory of aging holds tremendous promise, but is clearly a part of a bigger story, and integration into the existing paradigms should be considered at the outset. For example, autophagy can be included into the metabolism theory of aging, which includes metabolism of metabolites, reactive species, proteins, and organelles. Another example is the hormesis theory, which suggests that weak stress induces adaptive changes in the organism and may correlate with longevity30,48,55,82,83, thus, as Nietzsche opined, what does not kill you may make you stronger.\n\n\nAging-associated predictors of chronic disease\n\nExpansion of adipose tissue during positive energy balance is a primary determinant of obesity and related pathophysiological sequalae84,85. In contrast to the long-held belief that pro-inflammatory signals exert a negative impact on metabolism, emerging studies suggest that proinflammatory signaling is required for adipose tissue remodeling and expansion85. In fact, impairments in the pro-inflammatory response of adipose tissue were shown to increase ectopic lipid accumulation, glucose intolerance and systemic inflammation. In a related fashion, adipose tissue acquires a pro-inflammatory and senescent-like state with aging independent of adiposity86. Although macrophage infiltration is thought to be responsible for much of the increase in inflammation with obesity, infiltration is lower with aging in the absence of obesity suggesting that pro-inflammatory responses to aging may be mediated to reduce ectopic lipid storage. Possible evidence for this, described in a recent review86, indicates that the pro-inflammatory cascade with aging in adipose tissue is mediated through preadipocytes. It is possible that progenitor turnover, higher fatty acid levels, and toxic metabolites may contribute to fat cell senescence which then increases release of pro-inflammatory cytokines, increasing the potential for senescence to spread locally from cell to cell. It is important to determine whether the pro-inflammatory cascade that occurs with aging is derived from an aging-induced attempt to sequester lipids in relevant adipose storage sites as opposed to deleterious storage in ectopic sites such as skeletal muscle and liver.\n\nAs described in previous sections, CR extends lifespan in rodents and controversially in non-human primates as well29,87. At least part of the mechanism by which CR extends lifespan is through reductions in all-cause mortality. In fact, prolonged CR lowers body fat, slows rate of muscle loss, and lowers the incidence of neoplasia, diabetes, and cardiovascular disease in rhesus monkeys29. In addition, reproductive endpoints and brain morphology are preserved with CR. Since CR has been initiated at an early age in most studies, future studies are warranted to understand the effects of CR when initiated in adulthood. It is particularly important to understand the effects of CR following prolonged positive energy balance in adult rodents and non-human primates. Since recent human studies have suggested that exercise training can ameliorate the negative consequences of weight gain, it would be particularly beneficial to observe the effects of exercise training (forced or voluntary) during periods of CR and weight gain. Observation of mitochondrial function during these periods of CR and caloric surplus combined with either exercise training or no training would also be of great value.\n\nDietary restriction of the essential amino acid methionine has been shown to increase lifespan in rodents38 suggesting that restriction of total calories is not an absolute requirement for extending lifespan. While restricting fat and carbohydrate intake produces no extension of lifespan or improvements in markers of aging, improvements in the quality of carbohydrate consumption may be involved. Along these lines, a lower glycemic index (GI) diet can attenuate blood glucose excursions, lower postprandial lipemia, and decrease inflammatory markers. This is critical as many of these processes have been linked to age-related macular degeneration (AMD) and other aging associated diseases. Recent work by Uchiki et al.88, using mouse and cell models, revealed that consumption of higher GI diets may increase risk for AMD and other aging diseases by increasing glycative stress-associated lesions in several different tissue compartments. The increase in advanced glycation end products (AGE) were due to a decrease in both the ubiquitin-proteasome system and lysosome/autophagy pathway, which are critical for degrading AGEs. Thus, it appears that one potential mechanism by which a higher GI diet can increase risk for AMD and other diseases is by altering cellular proteolytic activity. Future studies should be conducted to elucidate additional biological mechanisms by which consumption of a lower GI diet can promote health.\n\n\nUse and effects of mesenchymal stem cells (MSCs)\n\nSignificant changes in fat mass and distribution occur during aging and the redistribution of adipocytes in the elderly has been linked with a number of age-related disorders. Notably, about 15–50% of the cells in adipocyte tissue comprise preadipocytes that give rise to fat cells89 and that play major roles in the mass and distribution of adipocyte cells during aging. Preadipocytes arise from adult stem cells in mammalian organs and one of the most important types of adult stem cell is the MSC. These are multipotent stromal cells that can differentiate into not only adipocytes, but also osteoblasts and chondrocytes as well as other cell types. Although MSCs have a high capacity for self-renewal, their ability to proliferate and differentiate decrease during aging, and with increasing passage number in vitro90,91. Considerable interest has focused on the mechanisms of MSC proliferation and differentiation that may provide novel approaches for tissue regeneration, a cornerstone of advances in interventional aging research.\n\nRecent work by Vidal et al.91 provides important insights into the senescence of MSCs that are central to organismal aging. To determine the potential use of MSCs in regenerative medicine, three different types of equine MSCs, bone marrow (BMSC), adipose tissue (ASC) and umbilical cord tissue (UCMSC), were evaluated for onset of cellular senescence in culture which has direct implications in their use for regenerative measures during aging, especially of fat and bone. It was discovered that BMSCs senesce after fewer population doublings than ASCs and UCMSCs demonstrating the more limited use of these types of MSCs for tissue regeneration. This also suggests that adipose and umbilical cord tissue may be preferable for tissue banking purposes although further studies on MSCs from different tissue sources are sorely needed in order to fully elucidate the potential for future therapeutic strategies in tissue repair during aging92. Potential differences in methodology, however, could account for some of the reported results of limited capacity of BMSCs for tissue regeneration and additional studies will be required for optimization and standardization of cell culture conditions for the different types of MSCs. Also, yet unsolved is the differentiation potential of the various MSCs (BMSCs, ASCs and UCMSCs) at higher passage, since loss of differentiation potential could severely limit their ability to create new tissue for regenerative medicine and their application in the elderly.\n\nA number of intriguing questions have arisen with respect to the mechanisms responsible for the attenuation of MSC proliferation and differentiation potential during aging. Among these remaining enigmas is whether senescence can occur at any stage of life86, which is clearly important with respect to the role of MSCs during aging. It is also not yet clear if senescence can spread from cell to cell in fat tissue in vivo or whether cellular senescence is a key cause for metabolic dysfunction secondary to age- and obesity-related changes in adipose tissue86. Future studies will also need to be directed toward determining if a senescence-like state can develop in terminally differentiated cells. Although there are suggestions that all of these venues are important with respect to MSCs, considerable additional research will be required before the role of MSCs in aging and their potential in tissue regeneration can be fully appreciated.\n\nAs aforementioned, MSCs are preadipocyte cells that give rise to new fat cells. The MSC progenitor cells can produce osteoblasts as well as macrophages that produce mesenchymal adipocyte-like default (MAD) cells important in age-related fat tissue redistribution and metabolic dysfunction. However, the variation in macrophage content during aging in subcutaneous fat tissue has shown little change while changes were noted between age and percent of macrophages in human omental adipose tissue92. This suggests that additional studies are required and that MSCs and subsequent macrophage distributions and abundance during aging among fat deposits will require considerable future investigation.\n\nLastly, MSCs are also central to bone marrow fat during aging and in age-related diseases such as diabetes. It is noteworthy that fat occupies a significant proportion of bone marrow; although its role in aging and age-associated disease is not well understood. A recent investigation by Krings et al.93 suggests that important changes occur in bone marrow fat not only during aging, but also in diabetes. MSCs give rise to different types of fat such as brown and white adipose tissue (BAT and WAT, respectively) and the metabolic phenotypes of bone marrow fat have characteristics of both types of adipose tissue. It was found that a decrease in BAT-like characteristics with aging and diabetes may contribute to age-related loss of bone remodeling and hematopoiesis that could have important implications in a number of changes that occur with aging secondary to the dynamics of MSCs in bone marrow93. This may create avenues for therapeutic interventions for bone integrity during aging and in age-associated diseases in addition to regenerative medicine possibilities for skeletal tissue. However, future studies will be required to determine the role of MSCs in bone marrow during aging and metabolic diseases especially with respect to their lineage identity and differential adipose content in bone marrow.\n\nTaken together, although a number of studies have eloquently revealed many of the roles of MSCs with respect to aging and age-related diseases, there are many avenues for additional studies to further reveal (1) the dynamics of cellular senescence in vivo and how this may limit potential for regenerative medicine approaches, (2) the behavior of MSCs from different tissue sources during aging, (3) the mechanisms for senescence of MSCs and how this impacts not only the aging process itself, but advances in intervention of aging and age-associated diseases, (4) the distribution and abundance of MSCs and macrophages during aging among fat depots, and (5) the role of MSCs in bone marrow and the importance of MSCs in contributing to maintenance of skeletal integrity in the elderly. Clearly many important advances have contributed significantly to our understanding of MSCs and their roles in aging; although many key questions remain that will provide fertile ground for future investigations.\n\n\nTelomeres relative to aging and energetics\n\nTelomeres, the repetitive sequences at the end of chromosomes that buffer the genetic coding regions, suffer from attrition during successive cycles of DNA replication and cellular division. The telomerase enzyme, responsible for the maintenance of telomere length, is thought to be active in germ-line cells, but relatively inactive in differentiated tissues in humans. Thus, the gradual erosion of the chromosome ends is proposed to contribute to genetic instability and loss under conditions of replicative stress and over the course of life94. When a critically short telomere length is reached, the cell enters replicative senescence and can no longer contribute to regenerative needs of the particular tissue95.\n\nTelomere length is variable between individuals at birth and with the proliferative demands of different body tissues, telomere length can vary between tissues within an individual94. Previous research has shown that shortened telomeres are associated with a number of metabolic- and age-related pathologies (e.g. oxidative stress, inflammation) and diseases (cardiovascular, diabetes, cancer, obesity)96–99. Controlling for individual variations in telomere length by using intra-individual telomere length, estimated attrition is further associated with potential contributors insulin resistance, a proposed mediator of metabolic-related disease and aging100.\n\nFrom a broader comparative perspective, humans have relatively short telomeres yet age more slowly and live significantly longer than other mammals. Despite these interspecies differences, the observed associations between telomere length and attrition suggest there may be additional information to be learned by understanding the physiologic contributors to telomere attrition that overlap with aging modulation. In this regard, a recent review discussing the obesity-mortality association has highlighted a number of gaps in the current knowledge99. For instance, there is a known association between excess body weight as measured by body mass index (BMI) and morbidity/mortality, yet there are exceptions where individuals with high-BMIs can have better than average health outcomes. Whether this BMI-mortality association is due to inter-individual differences in cellular aging rates, which may be captured through monitoring telomere change with body mass change, remains to be seen. Similarly, within individuals, whether telomere attrition is a composite outcome of age-sensitive biological mediators (oxidative stress, inflammation) that would reliably predict morbidity and mortality remains to be demonstrated. Along this same line, longitudinal studies of telomere dynamics within specific interventions to accelerate or retard aging (e.g., high fat diet feeding vs. dietary restriction), and including early life exposures and late-life health outcomes to better understand the “early life programming” effect in disease risk, could be informative. Such studies should include measures of caloric intake, energy balance, body composition, non-caloric nutritional factors associated with telomere length maintenance, stress exposure/protection, and co-morbidities in the outcomes. For the present, there is strong consensus that telomere shortening can contribute to cellular health and replicative capacity, and telomere length is correlated with a number of energetic- and age-associated negative health outcomes. Although it is not yet clear that telomere length or attrition is causative in these cases, particularly for mortality in humans, future research will require identifying the most relevant metric of telomere length (shortest single chromosome telomere length vs. population mean or overall distribution, etc.) and using such information to assess whether telomere attrition is more of a biomarker or causative factor in age-related phenotypes.\n\n\nAccretion and effects of body fat\n\nAging is generally associated with increases in total adiposity as well as waist circumference despite a lack of change or even a reduction in body weight (or concomitant decreases in lean mass)37,101. Positive energy balance, as resulted from subtle decreases in physical activity and basal metabolic rate that are not matched by decreases in energy intake, is suggested to be the cause for this increased adiposity101. During advanced old age, while total amount of fat mass tends to decline or remains stable, fat is redistributed from subcutaneous to intra-abdominal visceral depots and to ectopic sites, including muscle (cardiac and skeletal), liver and bone marrow86,89. Fat redistribution and fat tissue dysregulaton with aging occurs across species and is associated with age-related diseases, lipotoxicity, changed metabolic variables and reduced longevity, with some of these changes more closely related to regional adipose tissue distribution than total fat mass89,93. These changes might be explained by the increased systemic free fatty acid exposure caused by impaired capacity of fat tissue to store lipid89. Total or regional fat distribution with aging could have influence on morbidity and mortality risk, for example the removal of visceral fat significantly prolonged lifespan in rats101,102. In humans, weight gain and visceral adiposity is strongly associated with diabetes, atherosclerosis, thrombiosis, hypertension and other age-related diseases103. It is also notable that there appears to be protective effects of aerobic fitness on all-cause mortality, even at high levels of adiposity. Poor aerobic fitness is a powerful predictor of all-cause mortality across a wide BMI range104,105.\n\nBrown adipose tissue (BAT), which is related to thermal dysregulation and energy imbalance, decreases with aging in both animals and humans89. However, it is unclear whether the decline in BAT is related to the white adipose tissue changes with aging. With aging, midfacial fat compartments have an inferior migration and volume shift106.\n\nPreadipocytes, comprising a significant proportion of cells in adipose tissue, are constantly giving rise to new fat cells throughout life, but with aging they are less capable of accumulating lipid than are cells from younger individuals, which may contribute to the increased abundance of small, insulin-resistant, dysfunctional fat cells89. Preadipocytes also have decreased replication and adipogenesis, and increased proinflammatory cytokines and susceptibility to lipotoxicity with age86.\n\nFat, which was once thought of as a largely biologically inert lipid storage depot, also plays an important role in the regulation of energy metabolism, thermoregulation, inflammation, and immune responses that is implemented through a number of adipokines derived from adipose tissue2,93. Adiponectin, a cytokine primarily originating from adipose tissue, has been shown to negatively correlate with many age- and obesity-related diseases and positively correlate with longevity in mice107,108. Similarly, leptin, which has been shown to modulate total body fat and visceral fat distribution, might play a causative role in the metabolic decline in aging independent of fat mass109,110. As in obesity, aging is frequently associated with increased fat tissue and circulating pro-inflammatory cytokines secreted by dysfunctional fat cells, including tumor necrosis factors-α (TNF-α) and interleukin (IL-6)86, which in turn alter T lymphocyte subsets and attract mast cells and cause monocyte recruitment and macrophage activation89. Macrophages infiltrate adipose tissue of obese animals and humans more extensively in visceral fat than subcutaneous fat. Animal studies show macrophages and pro-inflammatory factors increase with aging mainly in subcutaneous fat but remain constantly high in visceral fat throughout life, which implicates that subcutaneous fat gets dysfunctional with aging, potentially contributing to fat redistribution86,89.\n\nUnder CR, proportionally more body fat is lost than other tissue, which may contribute to alterations in systemic metabolic homeostasis, influencing factors at the interface of metabolism and inflammation37. In aging studies utilizing non-human primates, CR lowers body weight, decreases fat mass and improves insulin sensitivity29,37. One of the most striking effects of CR in non-human primates is the impact on body composition, especially fat mass in the abdominal region29. A study on rats has shown that CR improves carbohydrate metabolism in aging by decreasing visceral fat111. Thus it is suggested that interventions that stimulate metabolism and/or activate WAT signaling hold great promise as mimics of CR37. However, it is not clear yet whether CR is as effective in non-overweight individuals or certain stages of lifespan. Furthermore, it is advisable to titrate allocations in already fat-depleted animals in order to guard against drawing on lean tissue for energy since food intake steadily declines after middle age29.\n\n\nThe aging heart\n\nAlthough other organs were discussed in the top 40 papers that we reviewed, the heart received particular emphasis. Congestive heart failure is an age-associated disease and is the direct result of defective cardiac aging112. The results from these studies suggest that fibrosis is a common consequence of myocardial infarction (MI) in older adults, but less common in younger adults113. The heart is a complex organ beating non-stop and is composed of 2 main cell types, cardiomyocytes and fibroblasts, along with pacemaker cells, Purkinje fibers, smooth muscle cells, resident macrophages, endothelial and epicardial cells, and a pool of resident stem cells. With aging, cardiomyocytes tend to either develop hypertrophy or wear and tear leading to necrosis. Fibroblast dysfunction with age leading to fibrosis, primarily contributes to diastolic heart failure in elderly individuals. Following a heart attack aging fibroblasts fail to develop stable scars and there is extensive left ventricle dilation and reduced survival in aging rodents and elderly individuals113,114. Novel strategies are required to reduce wear and tear in cardiomyocytes and preserve fibroblast function in order to form mature matrix in left ventricle remodeling in aging. Therapies are under investigation that stimulate cardiomyocyte regenerative capacity and limit fibrotic progress, but in the majority of cases, aging, the main confounding factor, is ignored.\n\nHypertension is highly associated with aging and congestive heart failure. Congestive heart failure significantly reduces cardiac output and limits functional capacity and quality of life. Age-related increases in the PR (PQ) interval (> 200 ms; distance on an ECG from the start of electrical conduction in the atria (beginning of p wave) to the start of electrical conduction in the ventricles (beginning of q wave) are positively associated with increased incidence of heart failure115. The relationship between ECG Pwave to Rwave interval (PRI) and heart failure has important clinical implications suggesting that monitoring of PRI may be a cost-effective strategy to identify individuals at increased risk for adverse outcomes associated with aging. Additional large-scale studies are needed to more precisely identify the magnitude of PRI prolongation over the lifespan. Future human studies should incorporate younger participants to elucidate the relationship between PRI and heart failure. In addition, an important area of investigation will be to determine if lifestyle interventions such as exercise and/or energy restriction (in the presence or absence of weight loss) reduces age-associated prolongation of PRI and subsequent incidence of heart failure and atrial fibrillation. Further validation with echocardiographic data such as chamber size and left ventricular mass are also crucial for clinical utility of the PRI in predicting adverse outcomes in multiple race/ethnicities.\n\nControl of inflammation is a widely studied approach for the treatment of MI, based on evidence from population-based studies. These studies suggest a direct role of inflammatory cytokines such as IL-6 and TNF-α on heart failure development116. Over the last thirty years, use of anti-inflammatory treatments has been discouraged for the treatment of post-MI inflammation in heart failure pathology. The reason for this is that studies with older animals indicate that aberrant inflammatory responses – either defective or overactive –are associated with worse outcomes, in particular reduced collagen deposition in response to injury. Interestingly, the reverse of what is observed in young animals. Therefore, an anti-fibrotic or anti-inflammatory strategy needs to be used with caution in older individuals. For example, methyl prednisolone treatment has been reported to cause ventricular rupture in humans and specific cyclooxygenase 2 inhibitors in fact induce more MI events. TNF-α trials too have so far failed to meet the expectation of cardiologist for heart failure treatment117–119. Thus, in the last three decades anti-inflammatory approaches have not been shown to be effective. Therefore, novel targets with a major focus on the resolution of inflammation are necessary.\n\nLindsey et al.120 showed that aging alters extracellular matrix remodeling and observed that middle aged (15 month-old) and older (23 month-old) CB6F1 mice had increased soluble protein fraction and decreased insoluble collagen compared to young mice. These structural changes are accompanied by increased end-diastolic dimension and ventricular wall thickness in aging mice with increased metalloproteinases (MMPs-3,8,9,12) and decreased inhibitory factors (TIMPs-3 and 4) expression. Thus, functional and proliferative capacites of fibroblasts are essential to maintain degraded matrix and to ensure that the structural integrity of the left ventricle is maintained in aging120.\n\nAge-related increases in fibrosis are also observed in C57Bl/6J sarcopenic and aging mice. Senescent (~32 month-old) C57BL/6J mice developed left ventricle structural changes characterized as interstitial fibrosis leading to increased left ventricular end diastolic dimension, decreased wall thickness, and decreased ejection fraction, indicating reduced contractile performance due to fibrosis.\n\nThus, collectively preventative strategies are required to prevent ventricular fibrosis in aging121. While developing prevention strategies, the post-MI therapeutic treatments are needed to maintain survival of cardiomyocte and preserve fibroblast function mainly focusing on resolution of inflammation rather than inhibition in the reparative phase of remodeling of the aging heart.\n\n\nMitochondria, reactive oxygen species, and cellular energetics\n\nDenham Harman originally proposed the “free radical theory of aging” in 1956, which was later refined to the “mitochondrial theory of aging” in 1970’s. The common theme in both was that oxidative damage accumulates with age; inclusion of the latter implicated reactive oxygen species generated by mitochondria as the primary source of this stress. This theory has been intensively explored – some studies have provided results consistent with this theory, whereas the majority has not. For example, in some cases oxidative stress increased lifespan, and other reports have shown that increased mitochondrial DNA (mtDNA) mutation associated with decreased function, without a concomitant increase in oxidative stress, is sufficient for generating a premature aging phenotype. At the inception of both these ideas the central paradigm for the free radical field was “oxidative stress” in which the simple concept was that the aging was accelerated by an imbalance between oxidants and antioxidants. The hypothesis was later refined to encompass data that suggested that mitochondrial superoxide and hydrogen peroxide were uncontrolled “leaks” from electron transfer and mitochondria were the major source of reactive species in the cell. Since the 1970’s it is clear that both concepts are not supported by experimental evidence and cannot encompass the role of reactive species such as hydrogen peroxide, nitric oxide and electrophilic lipids as cell signaling molecules71,122,123. Mitochondria are clearly not the major source of reactive species in the cell and whole families of proteins have now been discovered which show the controlled production of superoxide, nitric oxide, hydrogen peroxide and other reactive species with which the mitochondria interact in cell signaling cascades122,124–126. Consequently, mitochondrial function may prove to have a more critical role in aging, not because of its generation of damaging ROS, but because of its role in cell signaling.\n\nIt is instructive to consider the evolution of mitochondria, which are ancient bacterial symbionts with their own genetic and protein synthesis systems, to understand the fundamental nature of mitochondrial signaling to the cell, sometimes called retrograde signaling127. Each cell can contain hundreds to thousands of mitochondria each of which contains multiple copies of mtDNA. The mtDNA encodes genes essential for electron transport and oxidative phosphorylation. A unique aspect of mitochondria is that the proteins required for its multiple functions are encoded by both nuclear and mitochondrial genomes, requiring coordination and communication between the two cellular compartments. While reminiscent of its endosymbiotic origins, the mitochondrion retained many of the catalytic subunits required for electron transport and ATP synthase in its mtDNA. Hence, any mutation within the mtDNA has the potential to modify mitochondrial bioenergetics. In this respect, it has been proposed that prehistoric mutations in the mtDNA modulated the economics of mitochondrial production in the interrelated functions of ATP (energy), heat (thermoregulation), and superoxide (which can be converted to H2O2 and act as a signaling molecule) production, enabling our ancestors to successfully establish populations northward as they migrated from Africa. MtDNA mutations in northern migrants decreased caloric utilization for ATP generation while increasing that used for generating heat, creating a metabolic advantage for survival in colder climates. Although decreasing mitochondrial economy (in terms of ATP generation per calorie consumed), these changes were accommodated by changes in diet (increased caloric intake associated with animal fats). By comparison, mtDNA mutation that increased caloric utilization for ATP (thus decreasing energy lost to heat production) would be better adapted to low calorie diets and warmer climates (e.g., sub-Saharan Africa). In contemporary society, these prehistoric adaptations of mitochondrial economy can affect mitochondrial health and influence susceptibility to age-related disease in a setting of chronic positive energy balance72,128. In support of these concepts are molecular epidemiologic studies that correlate both human longevity and disease risk with certain mtDNA polymorphisms. More recently, it has been shown that mtDNA can directly influence susceptibility to heart failure and hepatic steatosis in mice, consistent with concepts of mitochondrial – nuclear genetic interactions129. Hence, the role of the mitochondrion in aging is most likely multi-faceted due to its multiple cellular functions that include bioenergetics and signaling. More recently, studies have implicated mtDNA as a DAMP (damage associated molecular pattern) broadening its cellular functions to include immune response. The focus in the redox biology field has now shifted to understand why and how the “quality” of the mitochondrial population in cells declines with age and integrates concepts from autophagy, cell signaling and bioenergetics71,130,131. Of particular interest is autophagy and mitochondrial biogenesis since they are both biological pathways that can link CR with healthy aging and bioenergetics37,132.\n\nHence, assessment of mitochondrial health in a dynamic fashion in our aging populations may provide a means for understanding the basis and progression of many forms of age-related diseases, including neurodegenerative diseases, diabetes, cardiovascular disease, liver disease, and cancer. This has now become possible with the development of measures of bioenergetic health from the leukocytes and platelets isolated from human subjects133,134.\n\n\nConclusion\n\nIn conclusion, across forty of the most cited papers at the interface of aging and energetics published in the last four years, we were able to identify ten major themes of research suggestions. We have summarized them here for the interested investigator.", "appendix": "Author contributions\n\n\n\nDBA conducted the search in SCOPUS.\n\nPB, JMG, TRN, MLP, DLS, MWS, TSS, TOT, SAW, and SNA initially each reviewed four of the 40 top A&E publications.\n\nDBA and LHA created thematic categories.\n\nSWB, VMD, GF, GVH, JLH, GRH, JLM, TRN, EPP, KAR, MWS, TSS, DLS, JDS, TOT, SWA, YY, JZ, and SNA condensed the initial reviews into summaries for each of the ten thematic categories.\n\nDBA, LHA, MBB, and SNA critically revised the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe project described was supported by Award Number P30DK056336 from the National Institute Of Diabetes And Digestive And Kidney Diseases and Award Number T32HL072757 from the National Heart, Lung, and Blood Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute Of Diabetes And Digestive And Kidney Diseases, the National Institutes of Health, or any other organization.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary materials\n\nAppendix A: Top ten most cited papers on aging and energetics in each of the four years that include 2010, 2011, 2012, and 2013.\n\nThe articles are organized in order of the number of citations within each year.\n\nClick here to access the file.\n\nhttp://dx.doi.org/10.5256/f1000research.5212.s35356\n\n\nReferences\n\nOsborne TB, Mendel LB, Ferry EL: The Effect of Retardation of Growth Upon the Breeding Period and Duration of Life of Rats. Science. 1917; 45(1160): 294–295. PubMed Abstract \| Publisher Full Text\n\nAnderson RM, Weindruch R: Metabolic reprogramming, caloric restriction and aging. Trends Endocrinol Metab. 2010; 21(3): 134–141. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrepanowski JF, Canale RE, Marshall KE, et al.: Impact of caloric and dietary restriction regimens on markers of health and longevity in humans and animals: a summary of available findings. Nutr J. 2011; 10: 107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSwindell WR: Dietary restriction in rats and mice: a meta-analysis and review of the evidence for genotype-dependent effects on lifespan. Ageing Res Rev. 2012; 11(2): 254–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRikke BA, Liao CY, McQueen MB, et al.: Genetic dissection of dietary restriction in mice supports the metabolic efficiency model of life extension. Exp Gerontol. 2010; 45(9): 691–701. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSutphin GL, Kaeberlein M: Dietary restriction by bacterial deprivation increases life span in wild-derived nematodes. Exp Gerontol. 2008; 43(3): 130–135. PubMed Abstract \| Publisher Full Text\n\nLiao CY, Rikke BA, Johnson TE, et al.: Fat maintenance is a predictor of the murine lifespan response to dietary restriction. Aging Cell. 2011; 10(4): 629–639. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLightfoot JT, Leamy L, Pomp D, et al.: Strain screen and haplotype association mapping of wheel running in inbred mouse strains. J Appl Physiol (1985). 2010; 109(3): 623–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLinden MA, Fletcher JA, Morris EM, et al.: Combining metformin and aerobic exercise training in the treatment of type 2 diabetes and NAFLD in OLETF rats. Am J Physiol Endocrinol Metab. 2014; 306(3): E300–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoberts MD, Company JM, Brown JD, et al.: Potential clinical translation of juvenile rodent inactivity models to study the onset of childhood obesity. Am J Physiol Regul Integr Comp Physiol. 2012; 303(3): R247–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark Y, Booth FW, Lee S, et al.: Physical activity opposes coronary vascular dysfunction induced during high fat feeding in mice. J Physiol. 2012; 590(Pt 17): 4255–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeindruch R, Walford RL: The retardation of aging and disease by dietary restriction. 1988. Reference Source\n\nMasoro EJ: Subfield history: caloric restriction, slowing aging, and extending life. Sci Aging Knowledge Environ. 2003; 2003(8): RE2. PubMed Abstract \| Publisher Full Text\n\nLibert S, Zwiener J, Chu X, et al.: Regulation of Drosophila life span by olfaction and food-derived odors. Science. 2007; 315(5815): 1133–7. PubMed Abstract \| Publisher Full Text\n\nHarper JM, Leathers CW, Austad SN: Does caloric restriction extend life in wild mice? Aging Cell. 2006; 5(6): 441–449. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarey JR, Liedo P, Harshman L, et al.: Life history response of Mediterranean fruit flies to dietary restriction. Aging Cell. 2002; 1(2): 140–148. PubMed Abstract \| Publisher Full Text\n\nLiao CY, Rikke BA, Johnson TE, et al.: Genetic variation in the murine lifespan response to dietary restriction: from life extension to life shortening. Aging Cell. 2010; 9(1): 92–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchleit J, Johnson SC, Bennett CF, et al.: Molecular mechanisms underlying genotype-dependent responses to dietary restriction. Aging Cell. 2013; 12(6): 1050–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrandison RC, Piper MD, Partridge L: Amino-acid imbalance explains extension of lifespan by dietary restriction in Drosophila. Nature. 2009; 462(7276): 1061–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBruce KD, Hoxha S, Carvalho GB, et al.: High carbohydrate-low protein consumption maximizes Drosophila lifespan. Exp Gerontol. 2013; 48(10): 1129–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller RA, Buehner G, Chang Y, et al.: Methionine-deficient diet extends mouse lifespan, slows immune and lens aging, alters glucose, T4, IGF-I and insulin levels, and increases hepatocyte MIF levels and stress resistance. Aging Cell. 2005; 4(3): 119–125. PubMed Abstract \| Publisher Full Text\n\nSolon-Biet SM, McMahon AC, Ballard JW, et al.: The ratio of macronutrients, not caloric intake, dictates cardiometabolic health, aging, and longevity in ad libitum-fed mice. Cell Metab. 2014; 19(3): 418–30. PubMed Abstract \| Publisher Full Text\n\nBarzilai N, Bartke A: Biological approaches to mechanistically understand the healthy life span extension achieved by calorie restriction and modulation of hormones. J Gerontol A Biol Sci Med Sci. 2009; 64(2): 187–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGreer E, Brunet A: The genetic network of life-span extension by dietary restriction. Handbook of the Biology of Aging. 2011; 3–23. Publisher Full Text\n\nGallinetti J, Harputlugil E, Mitchell JR: Amino acid sensing in dietary-restriction-mediated longevity: roles of signal-transducing kinases GCN2 and TOR. Biochem J. 2013; 449(1): 1–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKapahi P, Chen D, Rogers AN, et al.: With TOR, less is more: a key role for the conserved nutrient-sensing TOR pathway in aging. Cell Metab. 2010; 11(6): 453–465. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStenesen D, Suh JM, Seo J, et al.: Adenosine nucleotide biosynthesis and AMPK regulate adult life span and mediate the longevity benefit of caloric restriction in flies. Cell Metab. 2013; 17(1): 101–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBowman GL, Silbert LC, Howieson D, et al.: Nutrient biomarker patterns, cognitive function, and MRI measures of brain aging. Neurology. 2012; 78(4): 241–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKemnitz JW: Calorie restriction and aging in nonhuman primates. ILAR J. 2011; 52(1): 66–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmeisser S, Priebe S, Groth M, et al.: Neuronal ROS signaling rather than AMPK/sirtuin-mediated energy sensing links dietary restriction to lifespan extension. Mol Metab. 2013; 2(2): 92–102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorselli E, Maiuri MC, Markaki M, et al.: Caloric restriction and resveratrol promote longevity through the Sirtuin-1-dependent induction of autophagy. Cell Death Dis. 2010; 1: e10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWohlgemuth SE, Seo AY, Marzetti E, et al.: Skeletal muscle autophagy and apoptosis during aging: effects of calorie restriction and life-long exercise. Exp Gerontol. 2010; 45(2): 138–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShinmura K, Tamaki K, Sano M, et al.: Caloric restriction primes mitochondria for ischemic stress by deacetylating specific mitochondrial proteins of the electron transport chain. Circ Res. 2011; 109(4): 396–406. PubMed Abstract \| Publisher Full Text\n\nDazert E, Hall MN: mTOR signaling in disease. Curr Opin Cell Biol. 2011; 23(6): 744–55. PubMed Abstract \| Publisher Full Text\n\nBlagosklonny MV: Calorie restriction: decelerating mTOR-driven aging from cells to organisms (including humans). Cell Cycle. 2010; 9(4): 683–688. PubMed Abstract \| Publisher Full Text\n\nLamming DW, Ye L, Katajisto P, et al.: Rapamycin-induced insulin resistance is mediated by mTORC2 loss and uncoupled from longevity. Science. 2012; 335(6076): 1638–1643. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnderson RM, Weindruch R: The caloric restriction paradigm: implications for healthy human aging. Am J Hum Biol. 2012; 24(2): 101–106. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrentreich N, Matias JR, DeFelice A, et al.: Low methionine ingestion by rats extends life span. J Nutr. 1993; 123(2): 269–74. PubMed Abstract\n\nSegall PE, Timiras PS, Walton JR: Low tryptophan diets delay reproductive aging. Mech Ageing Dev. 1983; 23(3–4): 245–52. PubMed Abstract \| Publisher Full Text\n\nHector KL, Lagisz M, Nakagawa S: The effect of resveratrol on longevity across species: a meta-analysis. Biol Lett. 2012; 8(5): 790–793. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrune T, Jung T, Merker K, et al.: Decreased proteolysis caused by protein aggregates, inclusion bodies, plaques, lipofuscin, ceroid, and ‘aggresomes’ during oxidative stress, aging, and disease. Int J Biochem Cell Biol. 2004; 36(12): 2519–2530. PubMed Abstract \| Publisher Full Text\n\nMattson MP, Magnus T: Ageing and neuronal vulnerability. Nat Rev Neurosci. 2006; 7(4): 278–294. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLin MT, Beal MF: Mitochondrial dysfunction and oxidative stress in neurodegenerative diseases. Nature. 2006; 443(7113): 787–795. PubMed Abstract \| Publisher Full Text\n\nFinkel T, Holbrook NJ: Oxidants, oxidative stress and the biology of ageing. Nature. 2000; 408(6809): 239–247. PubMed Abstract \| Publisher Full Text\n\nSalmon AB, Richardson A, Perez VI: Update on the oxidative stress theory of aging: does oxidative stress play a role in aging or healthy aging? Free Radic Biol Med. 2010; 48(5): 642–655. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoutkooper RH, Williams RW, Auwerx J: Metabolic networks of longevity. Cell. 2010; 142(1): 9–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoutkooper RH, Mouchiroud L, Ryu D, et al.: Mitonuclear protein imbalance as a conserved longevity mechanism. Nature. 2013; 497(7450): 451–457. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYun J, Finkel T: Mitohormesis. Cell Metab. 2014; 19(5): 757–766. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang Y, Ikeno Y, Qi W, et al.: Mice deficient in both Mn superoxide dismutase and glutathione peroxidase-1 have increased oxidative damage and a greater incidence of pathology but no reduction in longevity. J Gerontol A Biol Sci Med Sci. 2009; 64(12): 1212–1220. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez VI, Bokov A, Van Remmen H, et al.: Is the oxidative stress theory of aging dead? Biochem Biophys Acta. 2009; 1790(10): 1005–1014. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLadiges W, Van Remmen H, Strong R, et al.: Lifespan extension in genetically modified mice. Aging Cell. 2009; 8(4): 346–352. PubMed Abstract \| Publisher Full Text\n\nPerez VI, Van Remmen H, Bokov A, et al.: The overexpression of major antioxidant enzymes does not extend the lifespan of mice. Aging Cell. 2009; 8(1): 73–75. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchriner SE, Linford NJ, Martin GM, et al.: Extension of murine life span by overexpression of catalase targeted to mitochondria. Science. 2005; 308(5730): 1909–1911. PubMed Abstract \| Publisher Full Text\n\nVan Remmen H, Ikeno Y, Hamilton M, et al.: Life-long reduction in MnSOD activity results in increased DNA damage and higher incidence of cancer but does not accelerate aging. Physiol Genomics. 2003; 16(1): 29–37. PubMed Abstract \| Publisher Full Text\n\nMesquita A, Weinberger M, Silva A, et al.: Caloric restriction or catalase inactivation extends yeast chronological lifespan by inducing H2O2 and superoxide dismutase activity. Proc Natl Acad Sci U S A. 2010; 107(34): 15123–15128. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakacs-Vellai K, Vellai T, Puoti A, et al.: Inactivation of the autophagy gene bec-1 triggers apoptotic cell death in C. elegans. Curr Biol. 2005; 15(16): 1513–1517. PubMed Abstract \| Publisher Full Text\n\nJuhasz G, Erdi B, Sass M, et al.: Atg7-dependent autophagy promotes neuronal health, stress tolerance, and longevity but is dispensable for metamorphosis in Drosophila. Genes Dev. 2007; 21(23): 3061–3066. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizushima N, Levine B, Cuervo AM, et al.: Autophagy fights disease through cellular self-digestion. Nature. 2008; 451(7182): 1069–1075. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nToth ML, Sigmond T, Borsos E, et al.: Longevity pathways converge on autophagy genes to regulate life span in Caenorhabditis elegans. Autophagy. 2008; 4(3): 330–338. PubMed Abstract\n\nVellai T, Takacs-Vellai K, Sass M, et al.: The regulation of aging: does autophagy underlie longevity? Trends Cell Biol. 2009; 19(10): 487–494. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVellai T: Autophagy genes and ageing. Cell Death Differ. 2009; 16(1): 94–102. PubMed Abstract \| Publisher Full Text\n\nMadeo F, Tavernarakis N, Kroemer G: Can autophagy promote longevity? Nat Cell Biol. 2010; 12(9): 842–846. PubMed Abstract \| Publisher Full Text\n\nRubinsztein DC, Marino G, Kroemer G: Autophagy and aging. Cell. 2011; 146(5): 682–695. PubMed Abstract \| Publisher Full Text\n\nMai S, Muster B, Bereiter-Hahn J, et al.: Autophagy proteins LC3B, ATG5 and ATG12 participate in quality control after mitochondrial damage and influence lifespan. Autophagy. 2012; 8(1): 47–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPyo JO, Yoo SM, Ahn HH, et al.: Overexpression of Atg5 in mice activates autophagy and extends lifespan. Nat Commun. 2013; 4: 2300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRubinsztein DC, Codogno P, Levine B: Autophagy modulation as a potential therapeutic target for diverse diseases. Nat Rev Drug Discov. 2012; 11(9): 709–730. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoseph AM, Adhihetty PJ, Wawrzyniak NR, et al.: Dysregulation of mitochondrial quality control processes contribute to sarcopenia in a mouse model of premature aging. PLoS One. 2013; 8(7): e69327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDodson M, Darley-Usmar V, Zhang J: Cellular metabolic and autophagic pathways: traffic control by redox signaling. Free Radic Biol Med. 2013; 63: 207–221. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee J, Giordano S, Zhang J: Autophagy, mitochondria and oxidative stress: cross-talk and redox signalling. Biochem J. 2012; 441(2): 523–540. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNixon RA: The role of autophagy in neurodegenerative disease. Nat Med. 2013; 19(8): 983–997. PubMed Abstract \| Publisher Full Text\n\nHill BG, Benavides GA, Lancaster JR Jr, et al.: Integration of cellular bioenergetics with mitochondrial quality control and autophagy. Biol Chem. 2012; 393(12): 1485–1512. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMitchell T, Chacko B, Ballinger SW, et al.: Convergent mechanisms for dysregulation of mitochondrial quality control in metabolic disease: implications for mitochondrial therapeutics. Biochem Soc Trans. 2013; 41(1): 127–133. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWu JJ, Quijano C, Chen E, et al.: Mitochondrial dysfunction and oxidative stress mediate the physiological impairment induced by the disruption of autophagy. Aging (Albany NY). 2009; 1(4): 425–437. PubMed Abstract \| Free Full Text\n\nGispert S, Ricciardi F, Kurz A, et al.: Parkinson phenotype in aged PINK1-deficient mice is accompanied by progressive mitochondrial dysfunction in absence of neurodegeneration. PLoS One. 2009; 4(6): e5777. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGautier CA, Kitada T, Shen J: Loss of PINK1 causes mitochondrial functional defects and increased sensitivity to oxidative stress. Proc Natl Acad Sci U S A. 2008; 105(32): 11364–11369. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarrison DE, Strong R, Sharp ZD, et al.: Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature. 2009; 460(7253): 392–395. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBjedov I, Partridge L: A longer and healthier life with TOR down-regulation: genetics and drugs. Biochem Soc Trans. 2011; 39(2): 460–465. PubMed Abstract \| Publisher Full Text\n\nZhang Y, Bokov A, Gelfond J, et al.: Rapamycin extends life and health in C57BL/6 mice. J Gerontol A Biol Sci Med Sci. 2014; 69(2): 119–130. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEisenberg T, Schroeder S, Andryushkova A, et al.: Nucleocytosolic depletion of the energy metabolite acetyl-coenzyme a stimulates autophagy and prolongs lifespan. Cell Metab. 2014; 19(3): 431–444. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRuckenstuhl C, Netzberger C, Entfellner I, et al.: Lifespan extension by methionine restriction requires autophagy-dependent vacuolar acidification. PLoS Genet. 2014; 10(5): e1004347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKramer PA, Ravi S, Chacko B, et al.: A review of the mitochondrial and glycolytic metabolism in human platelets and leukocytes: Implications for their use as bioenergetic biomarkers. Redox Biol. 2014; 2: 206–210. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalabrese V, Cornelius C, Dinkova-Kostova AT, et al.: Cellular stress responses, the hormesis paradigm, and vitagenes: novel targets for therapeutic intervention in neurodegenerative disorders. Antioxid Redox Signal. 2010; 13(11): 1763–1811. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLudovico P, Burhans WC: Reactive oxygen species, ageing and the hormesis police. FEMS Yeast Res. 2013. PubMed Abstract \| Publisher Full Text\n\nKim JY, van de Wall E, Laplante M, et al.: Obesity-associated improvements in metabolic profile through expansion of adipose tissue. J Clin Invest. 2007; 117(9): 2621–2637. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWernstedt Asterholm I, Tao C, Morley T, et al.: Adipocyte inflammation is essential for healthy adipose tissue expansion and remodeling. Cell Metab. 2014; 20(1): 103–1118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTchkonia T, Morbeck DE, von Zglinicki T, et al.: Fat tissue, aging, and cellular senescence. Aging Cell. 2010; 9(5): 667–684. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMattison JA, Roth GS, Beasley TM, et al.: Impact of caloric restriction on health and survival in rhesus monkeys from the NIA study. Nature. 2012; 489(7415): 318–321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUchiki T, Weikel KA, Jiao W, et al.: Glycation-altered proteolysis as a pathobiologic mechanism that links dietary glycemic index, aging, and age-related disease (in nondiabetics). Aging Cell. 2012; 11(1): 1–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSepe A, Tchkonia T, Thomou T, et al.: Aging and regional differences in fat cell progenitors - a mini-review. Gerontology. 2011; 57(1): 66–75. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJanzen V, Forkert R, Fleming HE, et al.: Stem-cell ageing modified by the cyclin-dependent kinase inhibitor p16INK4a. Nature. 2006; 443(7110): 421–6. PubMed Abstract \| Publisher Full Text\n\nVidal MA, Walker NJ, Napoli E, et al.: Evaluation of senescence in mesenchymal stem cells isolated from equine bone marrow, adipose tissue, and umbilical cord tissue. Stem Cells Dev. 2012; 21(2): 273–83. PubMed Abstract \| Publisher Full Text\n\nHarman-Boehm I, Blü her M, Redel H, et al.: Macrophage infiltration into omental versus subcutaneous fat across different populations: effect of regional adiposity and the comorbidities of obesity. J Clin Endocrinol Metab. 2007; 92(6): 2240–7. PubMed Abstract \| Publisher Full Text\n\nKrings A, Rahman S, Huang S, et al.: Bone marrow fat has brown adipose tissue characteristics, which are attenuated with aging and diabetes. Bone. 2012; 50(2): 546–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAviv A: The epidemiology of human telomeres: faults and promises. J Gerontol A Biol Sci Med Sci. 2008; 63(9): 979–83. PubMed Abstract\n\nVera E, Blasco MA: Beyond average: potential for measurement of short telomeres. Aging (Albany NY). 2012; 4(6): 379–92. PubMed Abstract \| Free Full Text\n\nO’Donovan, A, Pantell MS, Puterman E, et al.: Cumulative inflammatory load is associated with short leukocyte telomere length in the Health, Aging and Body Composition Study. PLoS One. 2011; 6(5): e19687. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilleit P, Willeit J, Mayr A, et al.: Telomere length and risk of incident cancer and cancer mortality. JAMA. 2010; 304(1): 69–75. PubMed Abstract \| Publisher Full Text\n\nFitzpatrick AL, Kronmal RA, Gardner JP, et al.: Leukocyte telomere length and cardiovascular disease in the cardiovascular health study. Am J Epidemiol. 2007; 165(1): 14–21. PubMed Abstract \| Publisher Full Text\n\nTzanetakou IP, Katsilambros NL, Benetos A, et al.: “Is Obesity Linked to Aging?”: adipose tissue and the role of telomeres. Ageing Res Rev. 2012; 11(2) : 220–29. PubMed Abstract \| Publisher Full Text\n\nGardner JP, Li S, Srinivasan SR, et al.: Rise in insulin resistance is associated with escalated telomere attrition. Circulation. 2005; 111(17): 2171–7. PubMed Abstract \| Publisher Full Text\n\nKuk JL, Saunders TJ, Davidson LE, et al.: Age-related changes in total and regional fat distribution. Ageing Res Rev. 2009; 8(4): 339–348. PubMed Abstract \| Publisher Full Text\n\nMuzumdar R, Allison DB, Huffman DM, et al.: Visceral adipose tissue modulates mammalian longevity. Aging cell. 2008; 7(3): 438–440. PubMed Abstract \| Publisher Full Text\n\nWilson P, Kannel W: Obesity, diabetes, and risk of cardiovascular disease in the elderly. Am J Geriatr Cardiol. 2002; 11(2): 119–123, 125. PubMed Abstract \| Publisher Full Text\n\nKodama S, Saito K, Tanaka S, et al.: Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA. 2009; 301(19): 2024–2035. PubMed Abstract \| Publisher Full Text\n\nWei M, Kampert JB, Barlow CE, et al.: Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA. 1999; 282(16): 1547–1553. PubMed Abstract \| Publisher Full Text\n\nGierloff M, Stohring C, Buder T, et al.: Aging changes of the midfacial fat compartments: a computed tomographic study. Plast Reconstr Surg. 2012; 129(1): 263–273. PubMed Abstract \| Publisher Full Text\n\nLubbers ER, List EO, Jara A, et al.: Adiponectin in mice with altered GH action: links to insulin sensitivity and longevity? J Endocrinol. 2013; 216(3): 363–374. PubMed Abstract \| Publisher Full Text\n\nYamauchi T, Kadowaki T: Adiponectin receptor as a key player in healthy longevity and obesity-related diseases. Cell Metab. 2013; 17(2): 185–196. PubMed Abstract \| Publisher Full Text\n\nMa XH, Muzumdar R, Yang XM, et al.: Aging is associated with resistance to effects of leptin on fat distribution and insulin action. J Gerontol A Biol Sci Med Sci. 2002; 57(6): B225–231. PubMed Abstract \| Publisher Full Text\n\nGabriely I, Ma XH, Yang XM, et al.: Leptin resistance during aging is independent of fat mass. Diabetes. 2002; 51(4): 1016–1021. PubMed Abstract \| Publisher Full Text\n\nBarzilai N, Banerjee S, Hawkins M, et al.: Caloric restriction reverses hepatic insulin resistance in aging rats by decreasing visceral fat. J Clin Invest. 1998; 101(7): 1353–1361. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLakatta EG: Age-associated cardiovascular changes in health: impact on cardiovascular disease in older persons. Heart Fail Rev. 2002; 7(1): 29–49. PubMed Abstract \| Publisher Full Text\n\nShih H, Lee B, Lee RJ, et al.: The aging heart and post-infarction left ventricular remodeling. J Am Coll Cardiol. 2011; 57(1): 9–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBujak M, Kweon H, Chatila K, et al.: Aging-related defects are associated with adverse cardiac remodeling in a mouse model of reperfused myocardial infarction. J Am Coll Cardiol. 2008; 51(14): 1384–1392. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMagnani JW, Wang N, Nelson KP, et al.: Electrocardiographic PR interval and adverse outcomes in older adults: the Health, Aging, and Body Composition study. Circ Arrhythm Electrophysiol. 2013; 6(1): 84–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalogeropoulos A, Georgiopoulou V, Psaty B, et al.: Inflammatory markers and incident heart failure risk in older adults: the Health ABC (Health, Aging, and Body Composition) study. J Am Coll Cardiol. 2010; 55(19): 2129–2137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoberts R, DeMello V, Sobel BE: Deleterious effects of methylprednisolone in patients with myocardial infarction. Circulation. 1976; 53(3 Suppl): I204–I206. PubMed Abstract\n\nMann DL, McMurray JJ, Packer M, et al.: Targeted anticytokine therapy in patients with chronic heart failure: results of the Randomized Etanercept Worldwide Evaluation (RENEWAL). Circulation. 2004; 109(13): 1594–1602. PubMed Abstract \| Publisher Full Text\n\nSellers RS, Radi ZA, Khan NK: Pathophysiology of cyclooxygenases in cardiovascular homeostasis. Vet Pathol. 2010; 47(4): 601–613. PubMed Abstract \| Publisher Full Text\n\nLindsey ML, Goshorn DK, Squires CE, et al.: Age-dependent changes in myocardial matrix metalloproteinase/tissue inhibitor of metalloproteinase profiles and fibroblast function. Cardiovasc Res. 2005; 66(2): 410–419. PubMed Abstract \| Publisher Full Text\n\nShioi T, Inuzuka Y: Aging as a substrate of heart failure. J Cardiol. 2012; 60(6): 423–428. PubMed Abstract \| Publisher Full Text\n\nLevonen AL, Hill BG, Kansanen E, et al.: Redox regulation of antioxidants, autophagy, and the response to stress: implications for electrophile therapeutics. Free Radic Biol Med. 2014; 71: 196–207. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMurphy MP: How mitochondria produce reactive oxygen species. Biochem J. 2009; 417(1): 1–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoncada S, Higgs EA: The discovery of nitric oxide and its role in vascular biology. Br J Pharmacol. 2006; 147(Suppl 1): S193–201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nErusalimsky JD, Moncada S: Nitric oxide and mitochondrial signaling: from physiology to pathophysiology. Arterioscler Thromb Vasc Biol. 2007; 27(12): 2524–31. PubMed Abstract \| Publisher Full Text\n\nAl Ghouleh I, Khoo NK, Knaus UG, et al.: Oxidases and peroxidases in cardiovascular and lung disease: new concepts in reactive oxygen species signaling. Free Radic Biol Med. 2011; 51(7): 1271–88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButow RA, Avadhani NG: Mitochondrial signaling: the retrograde response. Mol Cell. 2004; 14(1): 1–15. PubMed Abstract \| Publisher Full Text\n\nDunham-Snary KJ, Ballinger SW: Mitochondrial genetics and obesity: evolutionary adaptation and contemporary disease susceptibility. Free Radic Biol Med. 2013; 65: 1229–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFetterman JL, Zelickson BR, Johnson LW, et al.: Mitochondrial genetic background modulates bioenergetics and susceptibility to acute cardiac volume overload. Biochem J. 2013; 455(2): 157–67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGiordano S, Darley-Usmar V, Zhang J: Autophagy as an essential cellular antioxidant pathway in neurodegenerative disease. Redox Biol. 2014; 2: 82–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeber TA, Reichert AS: Impaired quality control of mitochondria: aging from a new perspective. Exp Gerontol. 2010; 45(7–8): 503–11. PubMed Abstract \| Publisher Full Text\n\nCorbi G, Conti V, Scapagnini G, et al.: Role of sirtuins, calorie restriction and physical activity in aging. Front Biosci (Elite Ed). 2012; 4: 768–78. PubMed Abstract\n\nChacko BK, Ravi S, Benavides GA, et al.: The Bioenergetic Health Index: a new concept in mitochondrial translational research. Clin Sci (Lond). 2014; 127(6): 367–73. PubMed Abstract \| Publisher Full Text\n\nChacko BK, Kramer PA, Ravi S, et al.: Methods for defining distinct bioenergetic profiles in platelets, lymphocytes, monocytes, and neutrophils, and the oxidative burst from human blood. Lab Invest. 2013; 93(6): 690–700. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "6106", "date": "22 Sep 2014", "name": "Holly M. Brown-Borg", "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 manuscript describing top research priorities in the aging and energetics field based on ideas extracted from recent papers. The title and abstract are appropriate and represent the information presented in the body of the manuscript. The basis of the paper is atypical in that the data generated are not assay-based but literature-based. The design is again, different but reasonable based on the outcomes to be generated. The data generated are ranked to identify the most highly cited papers in the past 4 years (2009<2014) and then further narrowed to 40 highly cited papers. The authors identified the top ten areas related to aging and energetics focusing on recommendations within these papers as to where further work should be focused. These areas included the CR longevity response, nutrient effects beyond energy, autophagy, mitochondria, ROS and cellular energetics and age-associated predictors of chronic disease among others. The recommendations for continued work in the areas covered were directed, insightful, and addressed strengths and weaknesses of the published studies and the area in general. Overall, the manuscript was well-written and provided important direction to those in the field in a relatively unbiased manner.", "responses": [] }, { "id": "6110", "date": "24 Sep 2014", "name": "Derek M. 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\nHere, the authors summarize 40 of the most widely-cited papers on energetics and aging over the last 4 years by themes, including suggested priorities from those publications for future work. Overall, I think this article does a good job of highlighting and summarizing state-of-the-art aspects of several broad and complex topics. Discussion of the heterogeneity of the caloric restriction (CR) response was particularly warranted. The implementation and interpretation of CR studies requires careful consideration of among things, the intervention timing (early or later in life), diet quality, the control group (ad libitum or mildly restricted) as well as the dose. Indeed, some strains that respond negatively to 40% CR, may respond differently at lower levels of restriction, in part due to reasons alluded to by the authors when discussing the potential risk of further fat depletion in already lean animals or humans. The authors also highlight some important recent observations regarding the role of specific nutrients (methionine) and botanical compounds (resveratrol) on aging, that do not involve calories per se. Of note, this has been further supported by a recent study that did not make the list (Levine et al., 2014), linking protein intake with IGF-1 levels, cancer and mortality risk in younger, but not older humans.Specific cellular mechanisms of interest are also discussed, with a particular emphasis on mTOR signaling in modulating organismal aging, and its interplay with nutrients and hormones. Rapamycin studies have driven much of the interest in this area, but metformin, which is not discussed here, is frequently and safely used clinically to treat diabetes, and may act in part via similar pathways, and has been linked to less cardiovascular disease, cognitive decline, cancers and improved survival in humans (see Bannister et al., 2014). Overall, this is a well conceived and timely review of recent highly-cited papers linking aging and energetics that deserves revisiting in another 4 years.", "responses": [] }, { "id": "6336", "date": "07 Oct 2014", "name": "George M. Martin", "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 order to make conclusions about the current consensus on mechanisms involved in the very broad field of aging and energetics (broadly defined as the acquisition, storage and utilization of metabolizable energy by biological organisms), the authors have utilized a novel methodology for determining what researchers regard as the top 40 relevant publications over the period of 2011 through 2013, as ascertained via a broad search algorithm and ranked by the number of citations over that time period. Their analysis of these selected papers provide an interesting selection of some ten overlapping domains, including such ”old timers” as caloric restriction (now more accurately referred to as dietary restriction) and the new star of the field, mTOR signaling.While this approach is an important guidepost to what has been discovered and, more importantly, as revealed by their analysis, what more needs to be done, the authors are encouraged to add an interesting and potentially informative control consisting of the lower 40 most cited papers revealed by their search algorithm. Such an exercise has the potential to reveal highly novel ideas that, perhaps 10 or so years from now, could result in a new flowering of publications that might transform some segments of this exceedingly complicated family of nature-nature interactions.", "responses": [] } ]	1	https://f1000research.com/articles/3-219
https://f1000research.com/articles/3-152/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "WikiPathways App for Cytoscape: Making biological pathways amenable to network analysis and visualization", "authors": [ "Martina Kutmon", "Samad Lotia", "Chris T. Evelo", "Alexander R Pico", "Chris T. Evelo", "Alexander R Pico" ], "abstract": "In this paper we present the open-source WikiPathways app for Cytoscape (http://apps.cytoscape.org/apps/wikipathways) that can be used to import biological pathways for data visualization and network analysis. WikiPathways is an open, collaborative biological pathway database that provides fully annotated pathway diagrams for manual download or through web services. The WikiPathways app allows users to load pathways in two different views: as an annotated pathway ideal for data visualization and as a simple network to perform computational analysis. An example pathway and dataset are used to demonstrate the functionality of the WikiPathways app and how they can be combined and used together with other apps. More than 2000 downloads between its first release in August 2013 and the submission of the paper in May 2014 highlight the importance and adoption of the app in the network biology field.", "keywords": [ "Pathways are commonly used as models for understanding biological processes. WikiPathways1 is an open", "collaborative", "wiki-based website for the curation of biological pathways that are more than just images. WikiPathways provides easy-to-use drawing and annotation tools to capture identities", "relationships", "comments and literature references for each pathway element and interaction. Contributed pathways are displayed like articles at WikiPathways and can be downloaded manually or programmatically through web services. This opens the possibility for pathway information to be accessed by other software tools for data visualization", "computational analysis and the interpretation of large-scale experimental data." ], "content": "Introduction\n\nPathways are commonly used as models for understanding biological processes. WikiPathways1 is an open, collaborative, wiki-based website for the curation of biological pathways that are more than just images. WikiPathways provides easy-to-use drawing and annotation tools to capture identities, relationships, comments and literature references for each pathway element and interaction. Contributed pathways are displayed like articles at WikiPathways and can be downloaded manually or programmatically through web services. This opens the possibility for pathway information to be accessed by other software tools for data visualization, computational analysis and the interpretation of large-scale experimental data.\n\nUtilizing WikiPathways web services, we developed an app for Cytoscape2, a network visualization and analysis software platform. The app queries and imports pathways from WikiPathways within the Cytoscape environment. Cytoscape’s core concepts are networks (nodes and edges), tables (rows and columns) and styles, which map table values to the visual properties of networks. Cytoscape leverages a rich ecosystem of apps to provide additional domain-specific semantics and data types, as well as custom visualization and analysis capabilities. With the WikiPathways app, we implemented two ways to represent a pathway as a Cytoscape network. In the first way, pathways are loaded with the complete visual appearance of the original at WikiPathways, including graphical annotations and labels. Once in Cytoscape, experimental data can be loaded as tables and visually mapped onto these pathway-style networks to provide biological context. In the second way, pathways are loaded as simplified networks, focusing on the biological entities and their interactions without any of the graphical elements of the original pathway diagram. The basic network style is ideal for topological analyses, network merging and automatic layout.\n\nIn this paper we present the implementation and usage of the WikiPathways app for Cytoscape. By bringing pathways into Cytoscape using the WikiPathways app, it is possible to make full use of pathway models with custom visualizations and computational analyses.\n\n\nImplementation\n\nThe WikiPathways app was developed for Cytoscape 3, which introduced a completely new software architecture. The new architecture is built on top of Open Service Gateway Initiative (OSGi)3, a software framework of pluggable modules and services. To be able to take advantage of the new architecture (Cytoscape API version 3.0.0), the predecessor to the WikiPathways app, the GPML Plugin, had to be rewritten.\n\nThe WikiPathways app employs the new architecture of Cytoscape in two ways. First, the app exports a user interface that can query and import pathways from the WikiPathways web service. Thanks to the service architecture in Cytoscape, this interface is seamlessly incorporated into Cytoscape’s “Import from Public Databases” dialog. Second, the app provides an API for programmatic access to the WikiPathways web services and the GPML file importer. Other apps can use the API to make queries to the WikiPathways web services and import GPML files without having to bundle the WikiPathways app. When the WikiPathways app is loaded in Cytoscape, the app registers the implementation of its API with the OSGi module system. Other apps can then request the API implementation through OSGi.\n\nThe new architecture also posed new challenges that required us to innovate with respect to visual styles. The new architecture includes a revamped model to represent networks. This model decouples the network topology and table data from its visual style. Visual styles constitute Cytoscape’s view model. When a node or edge is created in the network model, its view object is only created after a triggering of an event. Cytoscape does this to avoid redrawing of the network canvas while an app is still in process of building the network. Indeed, as the WikiPathways app reads a GPML file, it creates a series of nodes and edges in a network to represent the pathway. During this process, the app needs to assign visual styles to the nodes and edges it creates. However, as new nodes and edges are being added to the network, their view objects do not exist yet, making it impossible to assign their visual styles. To address this issue, we created a class called DelayedVizProp that stores our desired visual styles for nodes and edges. Once the network has been fully built, the app tells Cytoscape to create the view objects for the new nodes and edges. After that, the app looks through the DelayedVizProp instances and assigns nodes and edges their desired visual style.\n\nThe app relies on the PathVisio core library4 to read GPML files. The PathVisio library is included in the app. In previous versions of Cytoscape, apps that included libraries often conflicted with each other. Users had to painstakingly uninstall conflicting apps for Cytoscape to become usable again. OSGi solves this problem by insulating Cytoscape modules and apps from each other. Due to OSGi’s architecture in Cytoscape 3, the integrated PathVisio library is hidden from other apps and modules in Cytoscape and cannot conflict with them.\n\nThe app also uses the Apache HTTP Client library to make HTTP requests to the WikiPathways REST server. We avoided the Java built-in HTTP client class (java.net.HttpURLConnection), which is used frequently in Cytoscape and other apps. This class does not support cancellation. Proper cancellation is important for a responsive user interface. Users behind an interrupted internet connection should be able to back out of a WikiPathways request and return to Cytoscape. Each HTTP request is wrapped in a task, a unit of work in Cytoscape. When the user clicks cancel during the task execution, the app terminates the underlying HTTP request by calling the abort method in the Apache HTTP Client library.\n\n\nResults\n\nThe WikiPathways app in Cytoscape provides convenient access to the community-curated collection of biological pathways at WikiPathways. The functionality of the app is demonstrated here using the human Cardiac Hypertrophic Response pathway from WikiPathways (http://wikipathways.org/instance/WP2795) combined with an unpublished RNA-seq dataset that reflects gene expression levels during differentiation of cardiac stem cells (see Dataset 1). The logFC from timepoint 6 hrs vs control is visualized on the pathway. The human Cardiac Hypertrophic Response pathway contains gene products and metabolites involved in the intracellular signal-transduction pathways that coordinate Cardiac Hypertrophic Response. As described above, the WikiPathways app allows users to load pathways in two different views, as an annotated pathway and as a simple network (see Figure 1 and Figure 2). The example dataset and pathway will be used to explain how both views can be used in Cytoscape.\n\nLogFC values are visualized as node fill color with a color gradient from blue over white to red. Significant measurements (adjusted p-value < 0.05) are highlighted with a green border color. Elements in the pathway without a measurement are colored grey.\n\n(A) The simple network does not contain graphical annotations of the pathway. (B) NetworkAnalyzer was used to visualize node degree and betweenness of the nodes in the network to identify important hub nodes. (C) The logFC of the example dataset is visualized as node fill color with a gradient (blue over white to red) and the adjusted p-value < 0.05 is highlighted with a green border color. (D) jActiveModules finds active subnetworks (highlighted in purple) that are affected by varying gene expression.\n\nWhen loaded as a pathway, the precise layout of elements is identical to its representation at WikiPathways. The graphical elements, like labels and shapes, are included in the model in Cytoscape. As a pathway diagram, the full representation of biological information is visually preserved, which is ideal for providing a meaningful context for data visualization. Figure 1 shows the Cardiac Hypertrophic Response pathway loaded as an annotated pathway in Cytoscape. The Entrez Gene identifiers in the pathway were mapped to Ensembl using another app called BridgeDb (http://apps.cytoscape.org/apps/bridgedb) to match the identifiers used in the example dataset. The cardiac stem cell tissue development expression data can then be loaded, integrated and visualized on the pathway nodes (c.f. Introduction to Cytoscape tutorial, http://opentutorials.cgl.ucsf.edu/index.php/Tutorial:Introduction\\_to\\_Cytoscape\\_3.1-part2).\n\nWhen loaded as a network, all graphical annotations are removed and redundant nodes in the pathway are merged into one unique node in the network. Groups and complex interactions are visualized as very small nodes and a forced directed layout is applied. As an abstracted network graph, the same molecular relationships in the pathways can be made available for network analysis and augmentation. Figure 2A shows the Cardiac Hypertrophic Response pathway loaded as a network in Cytoscape. This simple network structure enables researchers to use other Cytoscape features and apps to merge two pathways, apply different layouts to the network or extend the pathway, for example, with regulatory interactions (CyTargetLinker5, http://apps.cytoscape.org/apps/cytargetlinker). It also enables users to investigate the topology of the network, like calculating degree and betweenness of the nodes with Cytoscape’s built-in NetworkAnalyzer tool to identify important hub nodes, see Figure 2B. Cytoscape also allows the visualization of experimental data in the network, as in Figure 2C which shows the cardiac stem cell tissue development expression data. There are several apps available for Cytoscape that provide methods that use experimental data to cluster nodes in the network (clusterMaker2, http://apps.cytoscape.org/apps/clustermaker2) or find subregions in the network affected by varying gene expression (jActiveModules, http://apps.cytoscape.org/apps/jactivemodules) as highlighted in Figure 2D.\n\n\n\n\nConclusions\n\nIn this paper we presented the WikiPathways app for Cytoscape, which allows the import of biological pathways as curated diagrams or as basic node-and-edge networks into Cytoscape. As shown in some examples, the app enables users to make full use of the pathway models by performing computational analyses and custom visualizations based on experimental data and network topology.\n\n\nSoftware availability\n\nSoftware available from the Cytoscape App Store: http://apps.cytoscape.org/apps/wikipathways\n\nLatest source code: https://github.com/wikipathways/cytoscape-wikipathways-app\n\nSource code as at the time of publication: https://github.com/F1000Research/cytoscape-wikipathways-app\n\nArchived source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.102047\n\nLicense: Lesser GNU Public License 3.0: https://www.gnu.org/licenses/lgpl.html\n\n\nData availability\n\nF1000Research: Dataset 1. Dataset studying differentiation of cardiac stem cells, 10.5256/f1000research.4254.d284156", "appendix": "Author contributions\n\n\n\nMK, SL and ARP participated in the design of the described software. SL and MK implemented the WikiPathways app. MK, SL, CTE and ARP contributed to the development and writing of this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe would like to acknowledge funding from NIGMS R01-GM100039 and NIDA U54-DA036134 (ARP, SL). This work was (co)financed by the Netherlands Consortium for Systems Biology (NCSB) which is part of the Netherlands Genomics Initiative/Netherlands Organisation for Scientific Research (MK, CTE).\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 Thomas Kelder for his work on the predecessor GPML plugin. Unpublished cardiac stem cell RNA-seq dataset courtesy of Dr. Bruce Conklin at the Gladstone Institutes.\n\n\nSupplementary materials\n\n\n\n\nReferences\n\nKelder T, van Iersel MP, Hanspers K, et al.: WikiPathways: building research communities on biological pathways. Nucleic Acids Res. 2012; 40(Database issue): D1301–D1307. 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\nOSGi Alliance. Osgi service platform, release 3. IOS Press, Inc., 2003. Reference Source\n\nvan Iersel MP, Kelder T, Pico AR, et al.: Presenting and exploring biological pathways with PathVisio. BMC Bioinformatics. 2008; 9(1): 399. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKutmon M, Kelder T, Mandaviya P, et al.: CyTargetLinker: a cytoscape app to integrate regulatory interactions in network analysis. PLoS One. 2013; 8(12): e82160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKutmon M, Lotia S, Evelo CT, et al.: Dataset studying differentiation of cardiac stem cells. F1000Research. 2014. Data Source\n\nKutmon M, Lotia S, Evelo CT, et al.: F1000Research/cytoscape-wikipathways-app. ZENODO. 2014. Data Source" }	[ { "id": "5286", "date": "02 Jul 2014", "name": "Christian A Grove", "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 is well written, clear and to-the-point. The title is appropriate and the abstract provides sufficient clarity and detail. The descriptions of the development, usage, and purpose of the WikiPathways App for Cytoscape are sufficiently detailed and informative for a bioinformatician such as myself. I cannot speak to the sufficiency of the detail of code and programming descriptions, although it seems reasonably complete. I've tested the App as the authors have described and found it to be easy to install and use with the set of descriptions and instructions provided on the source website and in this paper. All datasets and source code are easily obtainable with the links provided by the authors.I have found one small typo in the network/pathway which I have manually changed at the WikiPathways.org site: the original network referred to the node of a gene product as \"PCL\", which according to the cited reference and the position in the network is surely meant to represent \"Phospholipase C\" which has the canonical abbreviation of \"PLC\", in the citation for the network and in numerous previous studies. Unless I have made this observation in error, I would recommend changing this in the network and associated figures.", "responses": [ { "c_id": "887", "date": "02 Jul 2014", "name": "Martina Summer-Kutmon", "role": "Reader Comment", "response": "Dear Dr. Grove,Thank you for your very fast review. The typo you found is indeed a typo. We will update all figures accordingly. Thank you for immediately fixing it on WikiPathways yourself (http://wikipathways.org/instance/WP2795). This nicely highlights the efficiency of a wiki-based pathway archive." } ] }, { "id": "5523", "date": "04 Aug 2014", "name": "Ted Goldstein", "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\nYou don't really make the point in the conclusions of the key findings of this excellent work which are:It is now possible to do something quickly that before was tedious and therefore would not be done by anyone, except the most determined and experienced Cytoscape user. It is now possible to do something that could not be done before because you needed to add new functionality.I think it would be clearer that these are the main results if you add the following to the results section:I suggest that you make a before and after video. The before video is the experience of using Cytoscape, manually downloading the data from Wikipathways, failing to see the the nodes because it does not have the DelayedVizProp class etc.Then the after video shows the full experience.The elapsed time of the before video is something like 10 minute while the after video is something like 1 or 2 minutes. This will clearly demonstrate the improvement in the user experience. Please report these time numbers in the text in the results section and possibly summarize the numbers in the abstract. Of course, also post these videos to both F1000Research and Youtube as they represent the primary data of your study.", "responses": [ { "c_id": "936", "date": "07 Aug 2014", "name": "Alexander Pico", "role": "Author Response", "response": "Dear Dr. Goldstein,Thank you for the enthusiastic review and suggestions. We have drafted a new conclusions paragraph to include in the next revision to the article, which should be posted very soon. Coincidentally, we were gearing up to prepare a series of introductory videos for WikiPathways, each under 10 minutes long. We will certainly have one covering the use of the WikiPathways app for Cytoscape and will link/post it here as well. Given limited time and resources, we might not get around to producing the \"before\" video, but we will describe the process for contrast. In fact, the \"before\" approach is not so clear. There are many routes to transforming an arbitrary XML like GPML into a input file for Cytoscape, ranging from the most basic SIF to the full xGMML formats. And for each format, one must make judgement calls and compromises, and ideally rely on someone with detailed knowledge of the source format. So, that approach could take anywhere from a day to a week of scripting depending on the desired accuracy and purpose. And, of course, it takes months of planning and programming (like we did) to get the general-purpose app result we arrived at.Thanks again for highlighting these points and for your suggestions!" } ] } ]	1	https://f1000research.com/articles/3-152
https://f1000research.com/articles/3-209/v2	11 Sep 14	{ "type": "Opinion Article", "title": "Soil-transmitted helminths are a serious but understudied health concern in South Africa, requiring immediate attention from the scientific community.", "authors": [ "Hannah O Ajoge", "Stephen O Olonitola", "David R Smith", "Hannah O Ajoge", "Stephen O Olonitola" ], "abstract": "Parasitic roundworms and flatworms, defined broadly as helminths, are estimated to infect over 1 billion people worldwide, and are particularly prevalent in developing, resource-strained communities. The consequences of these infections are immense and wide reaching, resulting in massive reductions in local and global economic productivity and contributing to millions of deaths per year. Helminth diseases can also reduce vaccine efficacy and heighten morbidity rates of other serious illnesses, including tuberculosis and HIV/AIDS. Here, we argue that South Africa, which has one the highest rates of both HIV/AIDS and helminth infections on earth, needs to allocate more energy and resources into studying soil-transmitted helminths. Using PubMed and GenBank statistics, we show that the current South African research output on soil-transmitted helminths pales in comparison to that on HIV and tuberculosis. Basic research on helminth biology as well as on the social and environmental effects associated with infections could greatly reduce the burden of HIV/AIDS, tuberculosis, and other major illnesses in South Africa and beyond. The onus is on scientists, funding agencies, and governing bodies to channel efforts into studies on soil-transmitted helminths.", "keywords": [ "Ascaris lumbricoides", "helminthiasis", "parasitic worms", "South African helminths", "Trichuris trichiura" ], "content": "Commentary\n\nOver a billion people in sub-Saharan Africa and developing regions of Asia and the Americas have helminthiasis—a devastating category of diseases caused by parasitic worms (helminths)1–3, such as roundworms (nematodes), tapeworms (cestodes), or flukes (trematodes). Parasitic helminths can infect the intestinal tract, urinary tract, and blood of humans, and other mammals, leading to serious illness and in some cases permanent disability or death. The most common types of helminthiasis are intestinal and caused by soil-transmitted roundworms, including Ascaris lumbricoides and Trichuris trichiura, and the hookworms Necator americanus and Ancylostoma duodenale3–6. Schistosomiasis and lymphatic filariasis, commonly known as bilharzia and elephantiasis, respectively, are other less common kinds of helminthiases3.\n\nMost people with helminthiasis live in marginalized, resource-limited communities. Sadly, school-aged children and women of childbearing age, including adolescent girls, are among the most susceptible to chronic helminth infections and are at a greater risk of morbidity than members of other groups4. Because of its insidious, chronic nature, helminthiasis often goes untreated, leading to far-reaching social and economic consequences. Those infected with helminths show decreased performance at school and work and are predisposed to permanent disability, cardiovascular disease, and premature death. Helminthiases can also reduce vaccine efficacy and accentuate the morbidity rates of other serious diseases, such as malaria, tuberculosis, and HIV/AIDS3,7–9. Indeed, a number of studies on African populations have shown that helminth infections increase morbidity rates in HIV/AIDS patients as a result of systemic immune activation8,10–12.\n\nReducing the number of helminthiasis infections should, therefore, decrease disease burden and pressure on the health care system, especially in countries with high rates of HIV/AIDS. South Africa, which has the highest prevalence of HIV in the world and the highest rate of HIV-associated tuberculosis has channeled significant resources into fighting both of these diseases, but devotes comparatively little money and effort into combating helminthiasis13,14. Moreover, of the estimated 6.1 million people living with HIV in South Africa as of 2012, 0.41 million were under the age of 15 (http://www.unaids.org/en/regionscountries/countries/southafrica/), and are, thus, highly susceptible to helminth infections.\n\nMass drug administration (MDA) programs, involving cooperation between government health departments and major pharmaceutical companies, provide low-cost or free drugs for the treatment of various diseases and are one of the most cost-effective global public health control measures in existence5. Moreover, many of the MDA programs are targeted at preschool- and school-aged children; however, children who do not attend school can be overlooked1. Currently, a major obstacle to implementing cost-effective controls for helminthiasis is the lack of knowledge on the geographical distribution of infection and co-infection3,15 — a problem that could be easily and quickly solved with some basic research initiatives into helminthiasis. Again, given recent evidence of helminths developing drug resistance1,16, an alternative to MDA becomes a necessity. Vaccine (which doesn’t presently exist) research and development may thus be the best long-term strategy for sustained control and treatment of helminthic infections5,17.\n\nTo gauge South African research output on HIV and tuberculosis versus that on soil-transmitted helminths, we explored PubMed publication numbers and GenBank deposition data (Table 1). In the past year (July 2013 to June 2014), more than 400 journal articles have been published on HIV in South Africa, most of which have at least one author who is affiliated with a South African research institute. These journal articles cut across diverse HIV research areas, including diagnosis, treatment, vaccine trials, co-infections, mother-to-child transmission, and drug resistance. In comparison, there are very few published studies on soil-transmitted helminths in South Africa (mainly in Western Cape and Kwazulu-Natal). Within the past twelve months, only three journal articles on soil-transmitted helminths in South Africa appeared in PubMed, one of which was lead-authored by a researcher from a Cameroonian institute18–20 (Table 1). What’s more, there are thousands of HIV genome sequences from South Africans deposited in GenBank but not a single nucleotide sequence exists from a South African soil-transmitted helminth (Table 1). Given these statistics it appears that there is an overall lack of research on helminthiasis, both within and outside of South Africa.\n\nahttp://www.unaids.org/en/regionscountries/countries/southafrica/\n\nbhttp://www.who.int/tb/country/data/profiles/en/\n\nchttp://www.who.int/neglected_diseases/preventive_chemotherapy/databank/en/index.html\n\ndStatistics based on keyword searches in PubMed and GenBank using “HIV”, “Mycobacterium tuberculosis”, “Ascaris lumbricoides”, “Trichuris trichiura”, “Necator americanus”, and “Ancylostoma duodenale.” In addition to “South Africa”, each keyword was used to retrieve data (using the ‘Advanced Search Builder’ option) from the respective databases. For PubMed search, “2013/07/01 to 2014/06/30” was indicated as the Publication date. Search results were edited manually to remove incorrect entries. The incorrect entries were mostly articles that matched the main search terms but were from other African countries.\n\nThe onus is on scientists, funding agencies, and governing bodies to channel efforts into studies on soil-transmitted helminths. Basic research on helminth evolution, cell biology, genetics, and diversity as well as on the social and environmental effects associated with infections could greatly reduce the burden of HIV/AIDS, tuberculosis, and other major illnesses in South Africa and beyond. Helminth research also represents the first logical step in achieving the 2013 World Health Assembly (WHA) resolution to eliminate soil-transmitted helminths by 20201,3.", "appendix": "Author contributions\n\n\n\nHOA and SOO conceived the study. HOA and DRS prepared the first draft 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\nHOA was supported (2012–2013) by the South African National Research Foundation Innovation Post-doctoral fellowship. DRS is supported by a Discovery Grant to from the Natural Sciences and Engineering Research Council (NSERC) of Canada.\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 Anirban Banerjee for careful reading and helpful feedback on the manuscript.\n\n\nReferences\n\nBarry MA, Simon GG, Mistry N, et al.: Global trends in neglected tropical disease control and elimination: impact on child health. Arch Dis Child. 2013; 98(8): 635–641. PubMed Abstract \| Publisher Full Text\n\nGazzinelli A, Correa-Oliveira R, Yang GJ, et al.: A research agenda for helminth diseases of humans: social ecology, environmental determinants, and health systems. PLoS Negl Trop Dis. 2012; 6(4): e1603. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLustigman S, Prichard RK, Gazzinelli A, et al.: A research agenda for helminth diseases of humans: the problem of helminthiases. PLoS Negl Trop Dis. 2012; 6(4): e1582. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrooker S, Kabatereine NB, Smith JL, et al.: An updated atlas of human helminth infections: the example of East Africa. Int J Health Geogr. 2009; 8: 42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrichard RK, Basanez MG, Boatin BA, et al.: A research agenda for helminth diseases of humans: intervention for control and elimination. PLoS Negl Trop Dis. 2012; 6(4): e1549. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTchuem Tchuente LA: Control of soil-transmitted helminths in sub-Saharan Africa: diagnosis, drug efficacy concerns and challenges. Acta Trop. 2011; 120(Suppl 1): S4–11. PubMed Abstract \| Publisher Full Text\n\nFincham JE, Markus MB, Adams VJ: Could control of soil-transmitted helminthic infection influence the HIV/AIDS pandemic. Acta Trop. 2003; 86(2–3): 315–333. PubMed Abstract \| Publisher Full Text\n\nChachage M, Podola L, Clowes P, et al.: Helminth-associated systemic immune activation and HIV co-receptor expression: response to albendazole/praziquantel treatment. PLoS Negl Trop Dis. 2014; 8(3): e2755. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoolani Y, Bukhman G, Hotez PJ: Neglected tropical diseases as hidden causes of cardiovascular disease. PLoS Negl Trop Dis. 2012; 6(6): e1499. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalinkovich A, Borkow G, Weisman Z, et al.: Increased CCR5 and CXCR4 expression in Ethiopians living in Israel: environmental and constitutive factors. Clin Immunol. 2001; 100(1): 107–117. PubMed Abstract \| Publisher Full Text\n\nKalinkovich A, Weisman Z, Greenberg Z, et al.: Decreased CD4 and increased CD8 counts with T cell activation is associated with chronic helminth infection. Clin Exp Immunol. 1998; 114(3): 414–421. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSecor WE, Shah A, Mwinzi PM, et al.: Increased density of human immunodeficiency virus type 1 coreceptors CCR5 and CXCR4 on the surfaces of CD4(+) T cells and monocytes of patients with Schistosoma mansoni infection. Infect Immun. 2003; 71(11): 6668–6671. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChurchyard GJ, Mametja LD, Mvusi L, et al.: Tuberculosis control in South Africa: successes, challenges and recommendations. S Afr Med J. 2014; 104(3 Suppl 1): 244–248. PubMed Abstract \| Publisher Full Text\n\nSimelela NP, Venter WD: A brief history of South Africa’s response to AIDS. S Afr Med J. 2014; 104(3 Suppl 1): 249–251. PubMed Abstract \| Publisher Full Text\n\nBrooker S, Clements AC, Bundy DA: Global epidemiology, ecology and control of soil-transmitted helminth infections. Adv Parasitol. 2006; 62: 221–261. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCantacessi C, Hofmann A, Pickering D, et al.: TIMPs of parasitic helminths - a large-scale analysis of high-throughput sequence datasets. Parasit Vectors. 2013; 6: 156. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nResearch priorities for helminth infections. World Health Organ Tech Rep Ser. 2012; (972): xv–xvii, 1–174, back cover. PubMed Abstract\n\nSumbele IU, Ngole VM, Ekosse GI: Influence of physico-chemistry and mineralogy on the occurrence of geohelminths in geophagic soils from selected communities in the Eastern Cape, South Africa, and their possible implication on human health. Int J Environ Health Res. 2014; 24(1): 18–30. PubMed Abstract \| Publisher Full Text\n\nKvalsvig J, Albonico M: Effects of geohelminth infections on neurological development. Handb Clin Neurol. 2013; 114: 369–379. PubMed Abstract \| Publisher Full Text\n\nNxasana N, Baba K, Bhat V, et al.: Prevalence of intestinal parasites in primary school children of mthatha, eastern cape province, South Africa. Ann Med Health Sci Res. 2013; 3(4): 511–516. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "6286", "date": "08 Oct 2014", "name": "James 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\nThe article makes a reasonable point around the thorny issue of research prioritisation - although I think conflating actual numbers of child HIV infections with children at risk of infection of STH is not an ideal comparison given relative differences in disease burden. Ideally, the bibliographic data would be supplemented with direct insights into how and why research is prioritised (by all parties) in a South African context as the data itself, while initially striking, does not provide any great insight into this. I do not think the publishing results are surprising for many reasons - funding, research interest, building of a critical mass of research in areas of HIV research, need for new HIV tools, MDA relatively effective for STH etc. Given South Africa's slow response to its HIV crisis, in particular, it is not surprising so much research is devoted to HIV. I think, that being the case, and the fact that this is an opinion article that the authors might have lengthened their discussion and reflected on the relative evidence/impact of the three diseases (few would argue that HIV or TB is a bigger societal burden, but perhaps the data/analytics we use to assess disease burden is flawed or skewed towards certain measures of burden?) There is a large literature on this and other issues regarding prioritisation (DALYs etc), in relation to NTDs, that ought to be referenced.Furthermore, the interactions between funders, scientists and funding bodies are complex and systemic and it is overly simplistic to lay the onus on the research ecosystem to prioritise other diseases (there are also many reasons for inertia in research re-orientation). One needs to reflect on how to lever and incentivise the system if one wants different priorities supported, and to reflect on the data that unpins that sort of decision making. It would be useful for the authors to elaborate their argument here, which would involve having a deeper sense of the politics and practices of the STH (and broader) research community in South Africa. Finally, while quite clearly not much research is going on with regards to STH in South Africa the authors do not really provide evidence that research should be re-prioritised. MDA coverage - which happens in South Africa - means that the disease burden is relatively low, whereas HIV and TB (to a lesser extent) remain extraordinarily high. Few would argue that research should not be focused on HIV, at issue might be the extent to which it is. It would have strengthened the authors' case to have cited policy/opinion from the NRF, Dept of Health etc that indicates that soil-tranmitted helminths are a problem that requires more research activity, and suggests obvious areas where research is lacking.Ultimately the paper raises an interesting point about research and health priorities in a relatively resource constrained context but does not extend far enough beyond that by reflecting on the nature of evidence, exploring why helminth research is not better supported, and reflecting on how actually to lever more research if indeed it is truly necessary. If the authors were to focus on tracing precisely why so little soil-transmitted helminth research goes on (by for example interviewing the small number of people who do work on the diseases as well as key actors like the NRF) there may well be an interesting full length article to be written about science and health policy in South Africa.", "responses": [] }, { "id": "7877", "date": "09 Mar 2015", "name": "Chikwe Ihekweazu", "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 makes an important case and raises important issues. However, I am not sure how much children out of school is a challenge in South Africa. Also - I think the authors should have acknowledged the incredibly difficult challenge of prioritization of resource allocation. Otherwise - I would approve this for indexing.", "responses": [] } ]	2	https://f1000research.com/articles/3-209
https://f1000research.com/articles/3-217/v1	10 Sep 14	{ "type": "Method Article", "title": "PREMONITION - Preprocessing motifs in protein structures for search acceleration", "authors": [ "Sandeep Chakraborty", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Ravindra Venkatramani", "Abhaya M. Dandekar", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Ravindra Venkatramani", "Abhaya M. Dandekar" ], "abstract": "The remarkable diversity in biological systems is rooted in the ability of the twenty naturally occurring amino acids to perform multifarious catalytic functions by creating unique structural scaffolds known as the active site. Finding such structrual motifs within the protein structure is a key aspect of many computational methods. The algorithm for obtaining combinations of motifs of a certain length, although polynomial in complexity, runs in non-trivial computer time. Also, the search space expands considerably if stereochemically equivalent residues are allowed to replace an amino acid in the motif. In the present work, we propose a method to precompile all possible motifs comprising of a set (n=4 in this case) of predefined amino acid residues from a protein structure that occur within a specified distance (R) of each other (PREMONITION). PREMONITION rolls a sphere of radius R along the protein fold centered at the C atom of each residue, and all possible motifs are extracted within this sphere. The number of residues that can occur within a sphere centered around a residue is bounded by physical constraints, thus setting an upper limit on the processing times. After such a pre-compilation step, the computational time required for querying a protein structure with multiple motifs is considerably reduced. Previously, we had proposed a computational method to estimate the promiscuity of proteins with known active site residues and 3D structure using a database of known active sites in proteins (CSA) by querying each protein with the active site motif of every other residue. The runtimes for such a comparison is reduced from days to hours using the PREMONITION methodology.", "keywords": [ "The rapid development of crystallization techniques has resulted in a deluge of proteins with known structures1. Most of the proteins are annotated using sequence alignment methods by a ‘guilt by association’ logic based on the sequence-to-structure-to-function paradigm2. However", "sequence alignment methods are not applicable in cases where similar functional groups are identically positioned in the active site of proteins with no sequence homology. The classic example of this phenomenon", "known as convergent evolution3", "4", "is the major families of serine proteases (chymotrypsin and subtilisin)", "where the active site is structurally and functionally identical", "though there is no global sequence or structural homology5. According to some studies", "about 42% of entries annotated as ‘unknown functions’ are true examples of proteins of unknown function6." ], "content": "Introduction\n\nThe rapid development of crystallization techniques has resulted in a deluge of proteins with known structures1. Most of the proteins are annotated using sequence alignment methods by a ‘guilt by association’ logic based on the sequence-to-structure-to-function paradigm2. However, sequence alignment methods are not applicable in cases where similar functional groups are identically positioned in the active site of proteins with no sequence homology. The classic example of this phenomenon, known as convergent evolution3,4, is the major families of serine proteases (chymotrypsin and subtilisin), where the active site is structurally and functionally identical, though there is no global sequence or structural homology5. According to some studies, about 42% of entries annotated as ‘unknown functions’ are true examples of proteins of unknown function6.\n\nStructure-based methods have evolved to detect such convergently evolved proteins7,8. The conservation of structural properties is the primary driving logic behind many of these identification methods, reviewed in detail previously9. There are essentially two categories of programs that find binding sites in proteins (binding sites are typically closely related to protein function). The first one requires a predefined set of amino acids (motifs) of a known enzymatic function to search for the same within the protein under investigation8,10–14. The second category automatically detects similarity in the side chain patterns to classify protein functionality7,15–18. We have demonstrated through several detailed examples, using a method (CLASP19) which falls in the first category, that such structural conservation necessitates the conservation of electrostatic properties in proteins with the same functionality19–22.\n\nA challenge emerging in these methods relates to the large fold space of known proteins, although the rate of increase of this space is gradually being saturated23. Efficient parallelization has allowed the ProBiS algorithm to compare a protein query against the PDB in minutes15. To date, the identification of motifs is a task executed on the fly and applied sequentially13,14. Thus, running multiple queries involves several invocations of the same program. Our aim is to amortize the processing times by a one-time precompilation of all possible motifs, pruned using rational distance constraints, which can be leveraged for future queries.\n\nA simplistic approach to obtain motifs is to enumerate all possible combinations from the sequence. Motifs that span across distances rarely seen in active sites can be pruned out using structural information. In the present work, we propose a method to precompile all possible motifs comprising of a set (n=4 in this case) of predefined amino acid residues from a protein structure that occur within a specified distance (R) of each other (PREMONITION - Preprocessing motifs in protein structures for search acceleration). We have estimated R from the known active site residues of ~500 proteins annotated in the CSA database (http://www.ebi.ac.uk/thornton-srv/databases/CSA/)24. PREMONITION rolls a sphere of radius R along the protein fold centered at the Cα atom of each residue, and all motifs are extracted within this sphere. The maximum number of residues that occurs within R Å of any residue in protein dataset is also computed. This sets an upper bound for the polynomial complexity of the PREMONITION algorithm, and run times for the precompilation are reasonable. After such a precompilation step, the computational time required for querying a protein structure with multiple motifs is reduced considerably.\n\nPreviously, we had proposed a computational method (PROMISE) to estimate the promiscuity of proteins with known active site residues and 3D structure using the CSA database25. It took less than a minute to query one protein with all 500 motifs using PREMONITION, a process that took almost a day when done sequentially. Such speed up enables querying a much larger set of proteins using a comprehensive set of ligands, as is required in drug screening procedures.\n\n\nMaterials and methods\n\nAlgorithm 1 details the steps in creating the PREMONITION database for a given protein. We enumerate the steps using a concrete example for trypsin (PDBid:1A0J). Given a radius of interaction SOAS, we first compute the set residues within SOAS for each residue Residuei. For example, let us assume we are processing residue D102. Equation 1 gives the set of residues that have at least one atom within SOAS=10 Å from the Cα of D102.\n\nϕNearestResiduesD102 = [D102, W237, M90, V227, M180, G38, R62, S37, D194...] (1)\n\nWe now take combinations of n=4 from this set, which does not necessarily include D102 (Equation 2).\n\nϕCombinationsD102 = [(W237, M90, D102, V227), (W237, M90, D194, M180), (G38, R62, S37, D102)...] (2)\n\nAfter sorting the each combination based on the single letter amino acid code we obtain a set of n tuples (Equation 3).\n\nϕSortedStringsD102 = [DMVW = (102, 90, 227, 237), DDQV = (194, 90, 180, 237), DGRS = (102, 38, 62, 37)...] (3)\n\nNow, we add (102,90,227,237) to the global tableofmatches for the key ‘DMVW’. Thus, as we process every residue in the protein, we merge all occurrences of ‘DMVW’ (Equation 4).\n\nϕAllMotifsDMVW = [(194.104.53.141), (102.90.227.237), (194.104.52.51), (194.104.53.51)...] (4)\n\nExtracting all motifs of ‘DMVW’ now consists of the trivial task of reading this set from the file on disk.\n\nAdaptive Poisson-Boltzmann Solver (APBS) and PDB2PQR packages were used to calculate the potential difference between the reactive atoms of the corresponding proteins26,27. The APBS parameters and electrostatic potential units were set as described previously in19. Protein structures were rendered by PyMol (http://www.pymol.org/). The proteins were superimposed based on the matching motifs using DECAAF28.\n\n\n\nInput: P1 : Reference protein\n\nInput: n: number of residues in the motif\n\nInput: SOAS: Radius of sphere centered around each residue\n\nbegin\n\n/* Table mapping string of amino acids of length n to list of indices /\n\ntableofmatches = ∅ ;\n\nϕca = Cα atoms of all residues ;\n\nforeach CAi in ϕca do\n\nϕNearestResidues = FindResiduesWithinDist(CAi, SOAS);\n\nϕCombinations = GetCombinationsof_n(ϕNearestResidues, n);\n\nϕSortedStrings = SortBasedOnAminoAcidName(ϕSortedCombinations);\n\nInsertInTable(ϕSortedStrings,tableofmatches);\n\nend\n\n/ Output in file /\n\nforeach string in tableofmatches do\n\nϕMotifsstring = GetMotifsforEachString(tableofmatches);\n\nPrintListofMatches(ϕMotifsstring);\n\nend\n\nend\n\n\nResults and discussion\n\nFirst, we estimated the minimal radius of a sphere that encompasses active sites found in proteins. CSA provides catalytic residue annotation for enzymes in the PDB and is available online24. The database consists of an original hand-annotated set extracted from the primary literature and a homologous set inferred by PSI-BLAST2. We chose ~500 proteins from the CSA database that are annotated from the literature (SI list.doc). We computed the size of the active site (SOAS) for these known active sites by finding the minimum radius centered around one residue that encompassed the other residues. Table 1 shows the pairwise distance for the four residues that comprise the active site in trypsin (PDBid:1A0J) - Asp102, Ser195, His57 and Ala56. It can be seen that a sphere of radius 6.9 Å centered around His57 (c) would include all other residues. Other radii required to encompass all other residues of the motif and centered at a different residue is larger than this value (Asp102 = 7.8 Å, Ser195 = 9 Å and Ala56 = 9 Å). Thus, the SOAS for this protein is 6.9 Å. Figure 1a shows the frequency distribution of the SOAS for the set of 500 proteins chosen (mean = 7.3 Å, standard deviation = 1.8 Å, min = 3.5 Å and max = 13 Å). 90% of proteins have an SOAS below 10 Å.\n\nThe motif consists of Asp102/OD1(a), Ser195/OG(b), His57/ND1(c) and Ala56/N(d). A sphere of radius 6.9 Å centered around His57 (c) would include all other active site residues. This is the minimal distance - the radius of any sphere needed to include all residues is more than this value.\n\nData is extracted from ~500 proteins from the CSA database which are annotated from literature. (a) The size of the active site (SOAS) computed using the minimum radius centered around one residue that encompasses the other residues. (b) Number of residues enclosed by the SOAS (Red=10 Å, blue=11 Å). A residue Rj is considered to be within the SOAS of another residue Ri if any atom of Rj falls within a sphere of radius SOAS centered around the Cα of Ri.\n\nNext, we estimated the number of residues that fall within the SOAS radius in different proteins. PREMONITION takes combinations of 4 residues from each of this set (size N), and is thus polynomial in complexity (O(N) = N4). The value of upper bound on N needs to be known to ensure that runtimes are reasonable.\n\nFigure 1b shows the probability distribution of the number of residues that lie within a SOAS of 10 or 11 Å) for all residues in proteins in our dataset. A residue Rj is considered to be within the SOAS of another residue Ri if any atom of Rj falls within a sphere of radius SOAS centered around the Cα of Ri. Out of a total of 174780 residues in these proteins, the maximum number of residues found within the SOAS of 10 Å is 58. Thus, the upper bound on the number of possible combinations for one residue is 4C58 = 424270, which is quite tractable (as can be seen from runtimes below). Note, that although the algorithm is polynomial in complexity, this number increases rapidly with increasing motif length, as well as the SOAS. For example, for a motif length of 5, the number of possible combinations in a set of 58 residues is 4582116 (ten times the number for a motif length of 4). However, a 4 residue motif is sufficient to represent most active site conformations, and for a preliminary search on extensive datasets. Similarly, for a SOAS of 11 Å we obtain the maximum number of residues as 70 - which results in 4C70 (= 916895) possible combinations (twice the number for a SOAS of 10 Å).\n\nWe queried trypsin (PDBid:1A0J) with all the 500 motifs using the modified CLASP algorithm using the PREMONITION database19. The best matches are shown in Table 2. As expected, the best matches are those with known serine catalytic triads. As an illustrative example, we chose a protein (thioesterase - PDBid:1THT) with no known relationship with trypsin. This protein has the active site motif - H241 S114 V136 W213. This corresponds to the string query ‘HSVW’ (note that the string is sorted). We extract all entries for this string (Equation 5), which are all possible occurrences of the structural motif ‘HSVW’ in the protein.\n\nAs expected, the best matches are those with known catalytic triads.\n\nϕMotifsHSVW = [(71, 26, 121, 141), (71, 26, 138, 141), (57, 214, 212, 215)...] (5)\n\nAll entries of ‘HSVW’ are compared using CLASP. Table 3 shows the electrostatic potential difference and spatial difference in each of the motifs in Equation 5. The best scoring motif is (H57ND1,S195OG,V213N,W215NE1). However, even the best motif has a relatively large RMSD (Figure 2). Thus, this is not a significant match.\n\nThis motif corresponds to the key “HSVW”. D = Pairwise distance in Å. PD = Pairwise potential difference. APBS writes out the electrostatic potential in dimensionless units of kT/e where k is Boltzmann’s constant, T is the temperature in K and e is the charge of an electron.\n\nThe proteins are superimposed based on the matching active site residues using DECAAF28. (a) The global superimposition does not show any significant homology. (b) The detailed residue configuration. Residues from trypsin are in red, and those in the thioesterase are in blue. His57 and His241 completely overlap and are in black.\n\nPreviously, we had proposed a computational method (PROMISE) to estimate the promiscuity of proteins with known active site residues and 3D structure using the CSA database25. PROMISE used each of the 500 proteins with known active site residues extracted from the CSA database to query every other protein in that set. This procedure required ~500500=250000 program calls of CLASP19, each of which took one minute on an average. Thus, the total time taken was a month on a parallel system25. Using PREMONITION, it took less than a minute to query one protein with all 500 motifs. Thus, it took less than a day to replicate the PROMISE results. The precompilation step of extracting all motifs takes approximately 15 minutes on average for a single protein. For the protein with the largest SOAS (PDBid:1GPJ - 13 Å), the precompilation took one hour. These are acceptable values for a one time processing. The disk space for one protein PREMONITION file (zipped) is 14MB on average (7GB for 500 proteins).", "appendix": "Author contributions\n\n\n\nSC wrote the computer programs. All authors analyzed the data, and contributed equally to the writing and subsequent revisions of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAMD wishes to acknowledge grant support from the California Department of Food and Agriculture PD/GWSS Board. BJ acknowledges financial support from Tata Institute of Fundamental Research (Department of Atomic Energy). Additionally, BJR is thankful to the Department of Science and Technology for the JC Bose Award Grant. BA acknowledges financial support from the Science Institute of the University of Iceland.\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\nBernstein FC, Koetzle TF, Williams GJ, et al.: The Protein Data Bank: a computer-based archival file for macromolecular structures. J Mol Biol. 1977; 112(3): 535–542. PubMed Abstract \| Publisher Full Text\n\nAltschul SF, Madden TL, Schaffer AA, et al.: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25(17): 3389–3402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGherardini PF, Wass MN, Helmer-Citterich M, et al.: Convergent evolution of enzyme active sites is not a rare phenomenon. J Mol Biol. 2007; 372(3): 817–845. PubMed Abstract \| Publisher Full Text\n\nDoolittle RF: Convergent evolution: the need to be explicit. Trends Biochem Sci. 1994; 19(1): 15–18. PubMed Abstract \| Publisher Full Text\n\nRawlings ND, Barrett AJ: Evolutionary families of peptidases. Biochem J. 1993; 290(Pt 1): 205–218. PubMed Abstract \| Free Full Text\n\nNadzirin N, Firdaus-Raih M: Proteins of Unknown Function in the Protein Data Bank (PDB): An Inventory of True Uncharacterized Proteins and Computational Tools for Their Analysis. Int J Mol Sci. 2012; 13(10): 12761–12772. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRussell RB: Detection of protein three-dimensional side-chain patterns: new examples of convergent evolution. J Mol Biol. 1998; 279(5): 1211–1227. PubMed Abstract \| Publisher Full Text\n\nKleywegt GJ: Recognition of spatial motifs in protein structures. J Mol Biol. 1999; 285(4): 1887–1897. PubMed Abstract \| Publisher Full Text\n\nKonc J, Janezic D: Binding site comparison for function prediction and pharmaceutical discovery. Curr Opin Struct Biol. 2014; 25: 34–39. PubMed Abstract \| Publisher Full Text\n\nDebret G, Martel A, Cuniasse P: RASMOT-3D PRO: a 3D motif search webserver. Nucleic Acids Res. 2009; 37(Web Server issue): W459–464. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShatsky M, Shulman-Peleg A, Nussinov R, et al.: The multiple common point set problem and its application to molecule binding pattern detection. J Comput Biol. 2006; 13(2): 407–428. PubMed Abstract \| Publisher Full Text\n\nBauer RA, Bourne PE, Formella A, et al.: Superimpose: a 3D structural superposition server. Nucleic Acids Res. 2008; 36(Web Server issue): W47–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoyal K, Mohanty D, Mande SC: PAR-3D: a server to predict protein active site residues. Nucleic Acids Res. 2007; 35(Web Server issue): W503–505. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKirshner DA, Nilmeier JP, Lightstone FC: Catalytic site identification--a web server to identify catalytic site structural matches throughout PDB. Nucleic Acids Res. 2013; 41(Web Server issue): W256–265. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKonc J, Janezic D: ProBiS algorithm for detection of structurally similar protein binding sites by local structural alignment. Bioinformatics. 2010; 26(6): 1160–1168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolm L, Kaariainen S, Rosenstrom P, et al.: Searching protein structure databases with DaliLite v.3. Bioinformatics. 2008; 24(23): 2780–2781. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAngaran S, Bock ME, Garutti C, et al.: MolLoc: a web tool for the local structural alignment of molecular surfaces. Nucleic Acids Res. 2009; 37(Web Server issue): W565–570. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShulman-Peleg A, Shatsky M, Nussinov R, et al.: MultiBind and MAPPIS: webservers for multiple alignment of protein 3D-binding sites and their interactions. Nucleic Acids Res. 2008; 36(Web server issue): W260– 264. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Minda R, Salaye L, et al.: Active site detection by spatial conformity and electrostatic analysis--unravelling a proteolytic function in shrimp alkaline phosphatase. PLoS One. 2011; 6(12): e28470. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Ásgeirsson B, Minda R, et al.: Inhibition of a cold-active alkaline phosphatase by imipenem revealed by in silico modeling of metallo-β-lactamase active sites. FEBS Lett. 2012; 586(20): 3710–3715. PubMed Abstract \| Publisher Full Text\n\nRendon-Ramirez A, Shukla M, Oda M, et al.: A computational module assembled from different protease family motifs identifies PI PLC from Bacillus cereus as a putative prolyl peptidase with a serine protease scaffold. PLoS One. 2013; 8(8): e70923. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rendon-Ramirez A, Ásgeirsson B, et al.: Dipeptidyl peptidase-iv inhibitors used in type-2 diabetes inhibit a phospholipase c: a case of promiscuous scaffolds in proteins [v1; ref status: approved 1, approved with reservations 1, http://f1000r.es/2hw]. F1000Research. 2013; 2: 286. Reference Source\n\nJaroszewski L, Li Z, Krishna SS, et al.: Exploration of uncharted regions of the protein universe. PLoS Biol. 2009; 7(8): e1000205. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPorter CT, Bartlett GJ, Thornton JM: The Catalytic Site Atlas: a resource of catalytic sites and residues identified in enzymes using structural data. Nucleic Acids Res. 2004; 32(Database): D129–133. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rao BJ: A measure of the promiscuity of proteins and characteristics of residues in the vicinity of the catalytic site that regulate promiscuity. PLoS One. 2012; 7(2): e32011. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaker NA, Sept D, Joseph S, et al.: Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98(18): 10037–10041. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDolinsky TJ, Nielsen JE, McCammon JA, et al.: PDB2PQR: an automated pipeline for the setup of Poisson-Boltzmann electrostatics calculations. Nucleic Acids Res. 2004; 32(Web Server issue): W665–667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S: An automated flow for directed evolution based on detection of promiscuous scaffolds using spatial and electrostatic properties of catalytic residues. PLoS One. 2012; 7(7): e40408. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "6329", "date": "06 Oct 2014", "name": "Stefano Ciurli", "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 manuscript from Chakraborty et al. reports an algorithm aimed to accelerate the search in protein’s active site data bases. The algorithm precompiles all the possible motifs comprising a set of n=4 amino acids.Major points:The choice of considering combination of n=4 from the set of residues within the SOAS distance is not sufficiently explained and should be discussed. The authors write: “However, a 4 residue motif is sufficient to represent most active site conformations, and for a preliminary search on extensive datasets.” This claim should be justified and compared with similar choices in other algorithms performing similar tasks. The part discussing the illustrative example on thioesterase should be rewritten. First, it is not clear what the authors intend to show. Secondly, the properties reported in Table 3 (potential and spatial congruence) are not introduced anywhere in the paper and no discussion is provided about them. Finally, the criteria for the selection of the best scoring motif is not explained. The authors should rewrite some sections of the discussion in order to make it less technical and accessible to a widest audience of readers.Minor points:There are some typos scattered in the text. The article language, style and clarity would benefit greatly from a good proofreading by an English native speaker, if possible. The first reference of the paper is pretty old (1977) and should be replaced with that of RCSB Protein Data Bank (Berman et al., 2000). In general, some of the references in the introduction are quite old. The authors should replace these citations with recent reviews on the same topics. The last paragraph of the introduction should be moved at the end of the discussion. The acronym of the “size of the active site” parameter (SOAS) is used before its definition. Algorithm 1 should be depicted using a flow chart. In Table 2, the authors should change the caption (it is a partial repetition of the text) and mark the proteins with the serine catalytic triad. The section “Runtimes and disk space” should be moved in the Materials and Methods section.Given the above considerations, I suggest indexing this paper after major revisions.", "responses": [] }, { "id": "7709", "date": "24 Mar 2015", "name": "Juliana Bernardes", "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 work proposes a simple procedure for accelerating the search for structural motifs. It pre-compiles all motifs of size n within a radius R from protein structures and use this motif table to detect faster matches between a query sequence and proteins with known structure. Major points:This procedure is a trivial step since motif searches are not performed sequentially. A motif search algorithm must pre-compile of possible motifs in order to make the search feasible. In my opinion, it is a detail of CLASP implementation and the method is insufficient to justify a full Method Article - maybe a short paper. Minor:The paper is confusing and English must be improved.", "responses": [] }, { "id": "7706", "date": "24 Mar 2015", "name": "Xavier Barril", "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 concur with all comments made by the first referee. Additionally, it would be necessary to demonstrate that the method provides sound results using a benchmark set, comparing the results obtained with the original CLASP implementation. As it is, it is impossible to judge if the gain in computational performance is (totally or partially) offset by a loss in predictive capacity.", "responses": [] } ]	1	https://f1000research.com/articles/3-217
https://f1000research.com/articles/3-216/v1	10 Sep 14	{ "type": "Research Note", "title": "Using Twitter to investigate opinions about multiple sclerosis treatments: a descriptive, exploratory study", "authors": [ "Sreeram V. Ramagopalan", "Radek Wasiak", "Andrew P. Cox", "Radek Wasiak", "Andrew P. Cox" ], "abstract": "Background: Multiple sclerosis (MS) is a common complex disorder, with new treatment options emerging each year. Social media is being increasingly used to investigate opinions about drugs, diseases and procedures. In this descriptive exploratory study, we sought to investigate opinions about currently available MS treatments.Methods: The Twitter resource Topsy was searched for tweets mentioning the following MS treatments: Aubagio, Avonex, Betaferon or Betaseron, Copaxone, Extavia, Gilenya, Lemtrada, Novantrone, Rebif, Tysabri and Tecfidera between 1 Jan 2006 to 31 Jul 2014. Tweets were normalised and sentiment analysis performed.Results: In total, there were 60037 unique tweets mentioning an MS treatment. About half of the tweets contained non-neutral sentiment. Mean sentiment scores were different for treatments ranging from -0.191to 0.282 when investigating all tweets. These differences in sentiment scores between treatments were statistically significant (P<0.001). Sentiment scores tended to be higher for oral MS treatments than injectable treatments.Conclusions: Many tweets about MS treatments have a non-neutral sentiment. The analysis of social media appears to be a potential avenue for exploring patient opinion about MS treatments.", "keywords": [ "The analysis of social media is becoming a powerful tool that is being used increasingly to answer research questions across numerous areas including disease spatio-temporal epidemiology and drug adverse events1–3. Key stake-holders in the pharmaceutical industry", "including patients", "physicians", "regulatory authorities and pharmaceutical companies", "are increasingly using web technologies such as social media", "blogs and forums to generate and access opinions and real-world evidence of potentially medically important issues. This content serves as an important source of on-line medical opinions", "information and sentiments relating to particular drugs and events. The underlying assumption is that with access to such information", "a patient will be able to make more informed decisions about drugs", "diseases", "procedures and health-care providers." ], "content": "Introduction\n\nThe analysis of social media is becoming a powerful tool that is being used increasingly to answer research questions across numerous areas including disease spatio-temporal epidemiology and drug adverse events1–3. Key stake-holders in the pharmaceutical industry, including patients, physicians, regulatory authorities and pharmaceutical companies, are increasingly using web technologies such as social media, blogs and forums to generate and access opinions and real-world evidence of potentially medically important issues. This content serves as an important source of on-line medical opinions, information and sentiments relating to particular drugs and events. The underlying assumption is that with access to such information, a patient will be able to make more informed decisions about drugs, diseases, procedures and health-care providers.\n\nMultiple sclerosis (MS) is a chronic, neurodegenerative autoimmune disorder of the central nervous system (CNS). With a prevalence of one per 800 in North America and Northern Europe, MS is the most common acquired neurological disorder in young adults4. About 85% of patients present initially with relapsing-remitting MS (RRMS), characterized by recurrent episodes of neurological dysfunction interspersed with periods of lack of apparent disease activity4.\n\nAt present, there are nine disease modifying therapies (DMTs) approved by the US Food and Drug Administration (FDA) and 10 DMTs approved by the European Medicines Agency (EMA) for the treatment of RRMS, with new treatment options emerging each year. Approved treatments include interferons (Avonex, Betaferon, Betaseron, Extavia, Rebif), glatiramer acetate (Copaxone), natalizumab (Tysabri), and, more recently, the oral treatments teriflunomide (Aubagio), fingolimod (Gilenya), and dimethyl fumarate (Tecfidera). In this study, we explored whether we could analyse social media to help gauge patient sentiment about treatments using MS as an example. We used the popular social media site Twitter (http://twitter.com) to explore the reporting of patient sentiment and emotions about MS treatments.\n\n\nMethods\n\nThe Twitter resource, Topsy (http://topsy.com/), which houses all tweets made since 2006, was searched for the following brand names of MS treatments: Aubagio, Avonex, Betaferon or Betaseron, Copaxone, Extavia, Gilenya, Lemtrada, Novantrone, Rebif, Tysabri and Tecfidera using a daily search-time window (i.e. searching for tweets made every day), and specifying the English language. Brand names were used as we thought this would be more likely to reflect patient tweets and further the generic name for some MS treatments are not specific MS treatments. All dates from 1 Jan 2006 to 31 Jul 2014 were searched. For days in which there were more than 1000 tweets satisfying the search criteria, an hourly search-time window was applied for that day, to enable all available tweets to be found (the resource limits searches to 1000 results).\n\nTweets were downloaded in Extensible Markup Language (XML) format from topsy.com using the application program interface, otterapi (https://code.google.com/p/otterapi/).\n\nTweets were subsequently filtered to generate datasets for analysis:\n\n1. A unique dataset was generated from the “highlight” data class; thus, removing all directly copied retweets. This was performed so that sentiment analysis could be performed on unique tweets and not bias analyses by having several copies of the same tweet.\n\nAll subsequent filtering was case-insensitive.\n\n2. The unique dataset from (1) was filtered to remove items relating to company share prices/stockmarket news.\n\nThis was achieved by removing all tweets that contained:\n\na) “market_jp”, “thestreet”, “rtebusiness”, “pharma”,or “pharmsales” in the “permalink” dataclass,\n\nor\n\nb) “bloomberg”, “forbes”, “dow jones”, “financial times”, “stockpickr”, “marketwatch”, “business:”, “profit”, “shares”, or “sec” in the “highlight” data class. This filter was performed as we wanted to identify patient opinion about MS treatments and not stock market related tweets. This filter did retain tweets containing company names, some of which were stock/share price related but some tweets containing company names were from patients.\n\n3. The dataset from (2) was further filtered to remove items that mentioned the manufacturing companies by name: tweets were removed if they contained any of the following:\n\n“novartis”, “elan”, “biogen”, “merck”, “bayer”, “genzyme”, “sanofi”, “teva”, or “serono”. This filter was stringent and removed the majority of stock/share related tweets, but also removed some patient tweets.\n\nBecause of the short nature of tweets, typographical errors, ad-hoc abbreviations, phonetic substitutions, ungrammatical structures and emoticons are common, causing problems for text processing tools. Tokenisation and normalisation to make better sense of the tweet texts was achieved using TwitIE (http://gate.ac.uk/sale/ranlp2013/twitie/twitie-ranlp2013.pdf?m=1). Normalisation did not remove or alter any of the drug names.\n\nTweets were grouped into sequential monthly time periods for sentiment analysis using the twitteR R package (https://github.com/geoffjentry/twitteR/) and Jeffrey Breen’s sentiment analysis code (https://github.com/jeffreybreen/twitter-sentiment-analysis-tutorial-201107; a tutorial can be found at: http://www.inside-r.org/howto/mining-twitter-airline-consumer-sentiment). Word frequency analysis in tweets was performed using TagCrowd (http://tagcrowd.com). TagCrowd uses language-specific lists of common words which are removed from analysis.\n\nUsing lists of 2006 positive and 4783 negative words (http://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html#lexicon), the sentiment score for any tweet is calculated as follows:\n\nSentiment score = number of positive words - number of negative words\n\nIf the sentiment score > 0, this means that the sentence has an overall 'positive opinion', if the sentiment score < 0, this means that the sentence has an overall 'negative opinion', if the sentiment score=0, then the sentence is considered to be a 'neutral opinion'. Sentiment scores were summed for all tweets for each MS treatment, and means calculated. Mean sentiment scores were compared across treatments using the Kruskal-Wallis test. Statistical analysis was performed using R version 3.1.1 and p values less 0.05 were considered significant.\n\n\nResults\n\nIn total, there were 60037 unique tweets mentioning an MS treatment. The number of tweets by month is shown in Figure 1. Tweets for Tysabri started the earliest (January 2008) and Aubagio the latest (February 2009). When removing tweets that included share/stock information there were 56708 unique tweets and when removing tweets that included share/stock information or company names there were 41690 unique tweets.\n\nThe number of tweets by treatment, overall and when removing tweets that included share/stock information and/or company names is shown in Table 1. Tysabri had the largest number of tweets (n=14542, all tweets; n=10984 after filtering for company names and stock/share tweets) and Novantrone had the lowest (n=110, all tweets; n=109 after filtering for company names and stock/share tweets), both before and after filtering.\n\nThe sentiment score analysis of all normalised tweets, normalised tweets excluding those that contained share/stock information and normalised tweets excluding those that contained share/stock information and company names are shown in Table 2, Table 3 and Table 4. About half of all tweets in all analyses had a neutral sentiment (43–61%, all tweet data; 45–57% after filtering for company names and stock/share tweet data). Tweets for drugs that contained sentiment were more likely to be positive sentiment, apart from tweets for Novantrone and Tysabri (23–33% for drugs apart from Novantrone (16%), all tweet data; 24–31% for drugs apart from Novantrone (17%) and Tysabri (28%), after filtering for company names and stock/share tweet data).\n\nSumming sentiment scores for all tweets showed positive overall sentiment scores for all drugs apart from Novantrone (all analyses) and Tysabri (only after filtering for company names and stock/share tweet data). Gilenya had the highest summed sentiment score in all analyses. Boxplots of sentiment scores of all normalised tweets, normalised tweets excluding those that contained share/stock information and normalised tweets excluding those that contained share/stock information and company names are shown in Figure 2, Figure 3 and Figure 4. The mean sentiment score ranged from -0.191 to 0.282 (all tweet data); and -0.193 to 0.247 (after filtering for company names and stock/share tweet data). Novantrone always had the lowest mean sentiment score. Tecfidera had the highest mean score in the all tweet data, and Aubagio had the highest mean score in the filtered for company names and stock/share tweet data. The mean sentiment scores were different in all analyses (P<0.001 in the all tweet data, filtered for stock/share tweet data and filtered for company names and stock/share tweet data).\n\nMost common words in tweets for treatments were investigated. Example word clouds for the 50 most common words (excluding commonly used English words and drug names) in all normalised tweets for Avonex, Rebif and Tysabri are shown in Figure 5, Figure 6 and Figure 7. Of note is the frequency of ‘flu’ and ‘injection’ in Avonex and Rebif tweets and ‘infusion’ and ‘pml’ in Tysabri tweets.\n\n\nDiscussion\n\nWe present here, to the best of our knowledge, the first analysis of social media for MS treatments. A significant proportion of tweets did contain non-neutral sentiment about MS treatments, and the distribution of sentiment score was different between treatments. Thus it appears that Twitter can be a potential resource to understand patient opinion about MS treatments. When looking at frequency of words, notably ‘flu’ and ‘injection’ were in the 50 most common words in tweets about Rebif and Avonex and ‘infusion’ and ‘pml’ in the 50 most common words in tweets about Tysabri. Flu-like symptoms are well known side-effects of the injectible treatments Rebif and Avonex and progressive multifocal leukoencephalopathy or ‘pml’, is a well-known risk for patients taking the intravenously infused Tysabri5. This provides some sort of face validity for our results reflecting real specific tweets about MS treatments.\n\nInterestingly, the oral MS treatments- Gilenya, Aubagio and Tecfidera had the highest mean sentiment scores and Gilenya had the highest summed sentiment score in all analysis. This may reflect a well known patient preference for oral therapies as compared to injectible treatments6,7. Further work is needed to explore tweets in detail to see if the higher mean sentiment scores are related to positive tweets about the fact that these drugs are to be taken orally.\n\nThere are a number of limitations to this study. We are using automated tools to assign sentiment to tweet content- these tools will not recognise the intricacies of human language e.g. the context of the tweet and sarcasm for example. Further, whilst we tried to normalise tweets, the diversity of twitter slang will mean that abbreviations may not be recognised. We may have underestimated the number of tweets as we used brand names to identify drugs. Any tweets using the generic name or shortened versions will be missed. Whilst we tried to focus on tweets from patients, it is inevitable that business related tweets will have been included in our analysis and some patient tweets lost during filtering. It is also possible that not all tweets were delivered to us by the Twitter interface, although that is not possible to verify.\n\nOur findings and any interpretation should be regarded as speculative and exploratory. The results represent what can be potentially done relatively quickly and easily using data from Twitter. More rigorous analytical methods can be applied for more specific questions (e.g. the analysis of adverse events). It is clear from this study that tweets are written about MS treatments and many of these have a non-neutral sentiment. Further work is needed to look at these tweets in detail to further understand patient opinion about MS treatments.", "appendix": "Author contributions\n\n\n\nSVR is the guarantor of the study. SVR performed the analysis. APC contributed to the analysis and interpretation of the data. SVR wrote the first draft and all authors contributed to subsequent drafts and the final paper.\n\n\nCompeting interests\n\n\n\nSVR, RW and APC are employees of Evidera.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBroniatowski DA, Paul MJ, Dredze M: National and local influenza surveillance through Twitter: an analysis of the 2012-2013 influenza epidemic. PLoS One. 2013; 8(12): e83672. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsuya A, Sugawara Y, Tanaka A, et al.: Do cancer patients tweet? Examining the twitter use of cancer patients in Japan. J Med Internet Res. 2014; 16(5): e137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFreifeld CC, Brownstein JS, Menone CM, et al.: Digital drug safety surveillance: monitoring pharmaceutical products in twitter. Drug Saf. 2014; 37(5): 343–350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamagopalan SV, Dobson R, Meier UC, et al.: Multiple sclerosis: risk factors, prodromes, and potential causal pathways. Lancet Neurol. 2010; 9(7): 727–739. PubMed Abstract \| Publisher Full Text\n\nTanasescu R, Ionete C, Chou IJ, et al.: Advances in the treatment of relapsing-remitting multiple sclerosis. Biomed J. 2014; 37(2): 41–49. PubMed Abstract \| Publisher Full Text\n\nDibonaventura MD, Wagner JS, Girman GJ, et al.: Multinational Internet-based survey of patient preference for newer oral or injectable Type 2 diabetes medication. Patient Prefer Adherence. 2010; 4: 397–406. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFallowfield L, Atkins L, Catt S, et al.: Patients’ preference for administration of endocrine treatments by injection or tablets: results from a study of women with breast cancer. Ann Oncol. 2006; 17(2): 205–210. PubMed Abstract \| Publisher Full Text" }	[ { "id": "6808", "date": "28 Nov 2014", "name": "Bruce V Taylor", "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 a very interesting report and feel that as the authors do that this method of analysis may provide a real world window into patient perceptions of treatment, efficacy, and side effects. It would be interesting to assess churn in the twitter responses, for example PML in Tysabri is a major issue to patients and it would be interesting to see how this has trended over time, as the period for which tweets were collected includes the period of time when PML was a major issue (newly described) and would have been a major source of negative sentiment. Similarly for the newer agents it would be interesting to track these findings over time from release date and see whether positivity or negativity changes with patient experience.This type of research will undoubtedly provide significant and important patient focused outcomes for future medication usage and tolerance studies. at considerable cost efficiency.Can the authors look at trending words in these tweets, for example PML in Tecfidera or hair loss for Aubagio?", "responses": [] }, { "id": "6809", "date": "09 Dec 2014", "name": "Michael Hutchinson", "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\nAn interesting paper and an introduction to the reviewer of the new methods of analysis of social media. The authors have emphasised the pains that they took to exclude tweets from company and financial sources. The difficulty is that one cannot be certain that what they are analysing is not contaminated by tweets stimulated in some way from particular pharmaceutical companies. There is, of course, a \"newness\" affect and this explains the frequency of the tweets for Aubagio and Tecfidera. If one excludes Novantrone and Tysabri, then the lowest sentiment scores are for the interferons. What I find remarkable in this report is that the highest mean sentiment score is for Aubagio, a drug which many neurologists would have less enthusiasm for, given problems in its use in patients of childbearing age. Interesting findings and nicely presented paper, but I am still concerned about the source data; are all these tweets from genuine patients?", "responses": [] } ]	1	https://f1000research.com/articles/3-216
https://f1000research.com/articles/3-215/v1	09 Sep 14	{ "type": "Research Article", "title": "Directed evolution induces tributyrin hydrolysis in a virulence factor of Xylella fastidiosa using a duplicated gene as a template", "authors": [ "Hossein Gouran", "Sandeep Chakraborty", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Abhaya M. Dandekar", "Hossein Gouran", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Abhaya M. Dandekar" ], "abstract": "Duplication of genes is one of the preferred ways for natural selection to add advantageous functionality to the genome without having to reinvent the wheel with respect to catalytic efficiency and protein stability. The duplicated secretory virulence factors of Xylella fastidiosa (LesA, LesB and LesC), implicated in Pierce's disease of grape and citrus variegated chlorosis of citrus species, epitomizes the positive selection pressures exerted on advantageous genes in such pathogens. A deeper insight into the evolution of these lipases/esterases is essential to develop resistance mechanisms in transgenic plants. Directed evolution, an attempt to accelerate the evolutionary steps in the laboratory, is inherently simple when targeted for loss of function. A bigger challenge is to specify mutations that endow a new function, such as a lost functionality in a duplicated gene. Previously, we have proposed a method for enumerating candidates for mutations intended to transfer the functionality of one protein into another related protein based on the spatial and electrostatic properties of the active site residues (DECAAF). In the current work, we present in vivo validation of DECAAF by inducing tributyrin hydrolysis in LesB based on the active site similarity to LesA. The structures of these proteins have been modeled using RaptorX based on the closely related LipA protein from Xanthomonas oryzae. These mutations replicate the spatial and electrostatic conformation of LesA in the modeled structure of the mutant LesB as well, providing in silico validation before proceeding to the laborious in vivo work. Such focused mutations allows one to dissect the relevance of the duplicated genes in finer detail as compared to gene knockouts, since they do not interfere with other moonlighting functions, protein expression levels or protein-protein interaction.", "keywords": [ "The seminal and visionary work by Ohno in 1970 emphasized the pivotal role played by gene duplication in evolution. Gene duplication provides natural selection with the underlying mechanism to add functionality and adaptability to the genome by reusing pre-existing efficient and stable protein folds to catalyze novel reactions1. The fate of duplicated genes - unchanged functionality", "pseudogenization", "subfunctionalization or neofunctionalization - is the focus of intense research enabled by the advancements in sequencing technologies2", "3. Differing opinions on function innovation have also been articulated4", "and substantiated with real time evolution experiments5." ], "content": "Introduction\n\nThe seminal and visionary work by Ohno in 1970 emphasized the pivotal role played by gene duplication in evolution. Gene duplication provides natural selection with the underlying mechanism to add functionality and adaptability to the genome by reusing pre-existing efficient and stable protein folds to catalyze novel reactions1. The fate of duplicated genes - unchanged functionality, pseudogenization, subfunctionalization or neofunctionalization - is the focus of intense research enabled by the advancements in sequencing technologies2,3. Differing opinions on function innovation have also been articulated4, and substantiated with real time evolution experiments5.\n\nGene duplication plays a key role in the evolution of virulence-associated genes6. Secreted lipases, one of the highly replicated genes in Candida albicans7, has been implicated in disease pathogenesis, both in humans8 and in plants9. A lipase/esterase (LipA) from X. oryzae (Xanthomonas orysae pv. oryzaeraises (Xoo)) that causes bacterial blight in rice and is conserved across the genus Xanthomonas, has been recently characterized10. LipA also has three homologs (LesA, LesB and LesC) in the Xylella fastidiosa (Xf) genome11.\n\nXf is a major source of concern for both economic12 and food security reasons13, being the causal agent for Pierce’s disease of grape (PD) and citrus variegated chlorosis (CVC) of citrus species14. The presence of three duplicated genes (LesA/B/C) closely related to LipA from X. oryzae raises certain intriguing questions. It is logical to assume that these genes serve different purposes, since it is unlikely that three genes with identical functions will be maintained in the genome. A deeper understanding of their respective roles in the phenotypic context is essential in order to develop novel strategies to counter their threat15–17. LesA and LesB have 93% identity - yet, LesA can hydrolyze tributyrin whereas LesB can not. In the current work, we aim to induce tributyrin hydrolysis in LesB using minimal mutations.\n\nThe desire to mimic and accelerate natural evolution has fueled interest in directed evolution experiments, which endow or enhance functionality in enzymes. There has been some pioneering work in applying de novo methods to obtain catalytic functions18–22. However, most methods start with a template protein having the desired activity, known active site residues and 3D structure23–25. Previously, we have established a computational method (CLASP) based on spatial and electrostatic properties for the detection of active sites26–29, and a methodology to quantify promiscuity in proteins30. We also explored the prospect of promiscuous active sites to serve as the starting point for directed evolution (DECAAF)31,32. DECAAF has been applied to the problem of identifying mutations in LesB based on the active site of LesA in order to endow LesB with tributyrin hydrolysis.\n\nSince the structures of LesA/B are not known, and they share significant sequence homology with LipA (whose structure is known: PDBid:3H2G10), we used RaptorX to model the LesA/B structures. We first verified that the electrostatic profile of LesA and LesB are different. The LesA and LesB structures were then superimposed, and residues within a radii of 6 Å (MUT1:three residues) and 8 Å (MUT2:eight residues, including the three residues in MUT1) from the residues of the catalytic triad in LesA were compared to those in LesB. The differing residues were identified as the set of mutations which would induce tributyrin hydrolysis in LesB. It was observed that MUT1 and MUT2 residues are in two different contiguous stretches in the protein. As a validation step, we modeled the mutated sequences of LesB using RaptorX, and analyzed the differences in their electrostatic profiles. We created two mutants for LesB: LesBMUT1 with three mutations, and LesBMUT2 with eight mutations. The mutations in LesBMUT1 replicated the electrostatic congruence in the MUT1 residues, but not in the MUT2 residues. Consequently, we expected LesBMUT2 to have tributyrin hydrolysis, but not LesBMUT1.\n\nWe tested for the activity of LesB wild type and LesBMUT1/LesBMUT2 proteins using two assays. In one we use agar plates containing emulsified tributyrin, which upon hydrolysis of tributyrin to glycerol and butyric acid makes a clear zone visible. The other assay is a fluorescent quantitative assay in which 4-methylumbelliferyl butyrate (4-MUB) is used as a fluorescent substrate. The tributyrin hydrolysis activity of wild type LesB and LesBMUT1/LesBMUT2 with suggested mutations was tested in vitro using heterologous expression of these proteins in E. coli. We were able to confirm the tributyrin hydrolysis activity of LesBMUT2 in vitro using the assays mentioned above, while LesBMUT1 showed no activity.\n\n\nResults and discussion\n\nThe sequence alignment for LipA10, LesA (PD1703), LesB (PD1702) and LesC (P D1211) is shown in Figure 1a (gene names from http://www.ncbi.nlm.nih.gov/gene/). The phylogenetic tree and the pairwise sequence identity and similarity (Supplementary Table 1) suggests that LipA is more closely related to LesA/B than to LesC (Figure 1b). For example, LesA and LesB have 90% identity (347 out of 387 residues are identical) and 93% similarity (361 out of 387 residues are similar). In lieu of these differences in LesB, it does not have the capability to hydrolyze tributyrin that LesA does.\n\n(a) Sequence alignment. (b) Cladogram generated from (a).\n\nThe structures for LesA/B/C have not been solved. However, since there is significant sequence homology of these proteins with LipA, we modeled the structures of these proteins using RaptorX33. RaptorX automatically chooses the best template (LesA - PDBid:3H2G, LesB and LesC - PDBid:3H2I)10. The structural superimposition of these proteins done using MUSTANG34 is shown in Figure 2a.\n\n(a) MUSTANG generated superimposition (LipA - red, LesA - green, LesB - blue, LesC - magenta). (b) STEEP generated superimposition, obtained by superimposing three atoms from the catalytic triad in the active site. It can be seen that MUSTANG generates a better overall superimposition, since STEEP tries to get a better superimposition of the atoms, rather than achieve a global superimposition32. (c) Superimposition of LesA and LesB (red and yellow, respectively), obtained by superimposing the catalytic triad (Asp336, Ser176 and His377 in LipA). The Asp overlaps completely (at the origin of the coordinate system) and is shown in black. Some of the different residues are shown in magenta and blue for LesA and LesB respectively. (d) Sequence alignment of active site residues (residues within a radii of 7 Å from the residues of the catalytic triad in LesA). Note that these residues are not in a one contiguous stretch.\n\nThe active sites of these proteins are conserved and have a serine catalytic triad (Asp336, Ser176 and His377 in LipA). Previously, we have proposed a method to suggest mutations in a protein in order to endow it with a specific catalytic function based on structural and electrostatic homology of residues in the active site to a known protein with the desired function31. We chose the Cβ atoms of the residues as the representative atom for each residue for an uniform comparison (only glycine lacks a Cβ atom). We have shown that although the reactive groups are different for amino acids, this difference is encapsulated in the backbone Cβ atoms35. Table 1 shows the spatial and electrostatic potential difference (EPD) congruence in the catalytic triad in these proteins. We applied transformations to align the catalytic triad (Figure 2b)31.\n\nIt can be seen that the serine catalytic triad is congruent in all proteins. Further, LesA and LesB have a different electrostatic profile when we consider another residue (Ser75/Arg75 respectively) - see pair ‘bd’ for instance. While the EPD in this pair is moderately positive in LesA (+38.5 EPD units), this is seen to change sign in LesB (-63.1 EPD units). The mutated structures (LesBMUT1 and LesBMUT2) modulates this EPD back to moderately positive again (+19 and +12 units in LesBMUT1 and LesBMUT2, respectively). D = Pairwise distance in Å. PD = Pairwise potential difference. See Methods section for units of potential.\n\nThis multiple superimposition of the proteins provided a single frame of reference for comparing the proteins LesA and LesB (Figure 2c). After the superimposition, we took residues within a radii of 7 Å from the residues of the catalytic triad in LesA. Now, for each of these residues we found the closest residue in LesB - noting that they are now superimposed. The different residues, that form the set which are to be mutated, are shown in Table 2 and Figure 2d. It can be seen that these residues lie within two contiguous stretches: a) R75, D76 and G77 (MUT1) which are about 6 Å away from the active site residues and b) G363, E364, F366, F367 and T368 (MUT2) that are about 8 Å away from the active site residues. We created a LesB mutant by mutating resides in (a) (LesBMUT1), and another (LesBMUT2) by mutating residues in both sets.\n\nTwo sets of mutations were studies here - LesBMUT1 and LesBMUT2. The LesBMUT1 mutations are within 6 Å of the active site residues, and located within a contiguous stretch. We further added the LesBMUT1 mutations to another set of residues that differ (8 Å away from the active site residues) to obtain LesBMUT2 mutations.\n\nAs a validation step, we modeled the structures of the mutated sequences in order to compare the change in the electrostatic profile. We first consider the catalytic triad and one of the mutated amino acid (Arg75 to Ser75 in LesB) in the LesBMUT1/LesBMUT2 structures (Table 1). It can be seen that the mutated active site is spatially similar to wild type. The same holds true for all structures discussed henceforth. The major change in the electrostatic profile can be seen for the pair ‘bd’ (Ser165CB/Arg75CB in LesB versus Ser165CB/Ser75CB in LesBMUT1/LesBMUT12). While the EPD in this pair is moderately positive in LesA (+38.5 EPD units), this is seen to change sign in LesB (-63.1 EPD units). The mutated structures (LesBMUT1/LesBMUT2) modulate this EPD back to moderately positive again (+12.4 EPD units).\n\nWhile it may appear from Table 1 that the three mutations (MUT1) in LesBMUT1 might suffice, it is necessary to compute the changes in the MUT2 residues. We chose one residue from the active site (Asp323) and three residues from the MUT2 set (E364, F366 and F367). Since Gly does not have a Cβ atom, it could not be used for comparison. There are certain differences in the electrostatic profile between LesBMUT1 and LesBMUT2. While the EPD in ‘bd’ is moderately positive in LesA (+8.2 EPD units), this changes sign in LesBMUT1 (-72.3 EPD units) (Table 3). However, this EPD modulates back to moderately positive again in LesBMUT2. Thus, it might be necessary to introduce all eight mutations (MUT1 + MUT2) in order to introduce tributyrin hydrolysis in LesB. This verification step is unique to our methodology, and infuses confidence in the chances of success with the in vivo mutations. We now made two mutants of LesB - LesBMUT1 having three mutations, and LesBMUT2 having eight mutations.\n\nWe chose one residue from the active site (Asp323) and three residues from the MUT2 set (E364, F366 and F367). Since Gly does not have a Cβ atom, it could not be used for comparison. We observed certain differences in the electrostatic profile between LesBMUT1 and LesBMUT2 - see pair ‘bd’ for instance. While the EPD in this pair is moderately positive in LesA (+8.2 EPD units), this is seen to change sign in LesBMUT1 (-72.3 EPD units). This EPD modulates back to moderately positive again in LesBMUT2. Thus, it might be necessary to introduce all eight mutations (MUT1 + MUT2) in order to introduce tributryn catalysis in LesB. D = Pairwise distance in Å. PD = Pairwise potential difference. See Methods section for units of potential.\n\nTo test the activity of LesBMUT1/LesBMUT2 in vitro, we expressed LesB and LesBMUT1/LesBMUT2 in E. coli using an expression vector to obtain high quantities of heterologous protein. The E. coli strain expressing LesBMUT2 streaked on tributyrin agar plates showed significant tributyrin hydrolysis activity compared to wild type LesB. The activity was visualized by a tributyrin hydrolysis zone on the agar plate containing emulsified tributyrin fat (Figure 3a). However, the LesBMUT1 showed no tributyrin hydrolysis under the same conditions.\n\n(a) Tributyrin agar plate assay. Tributyrin hydrolysis zone (clearing zone) is visible for E. coli strains expressing LesBMUT2 and not for LesB or empty vector (EV) strains. (b) A4-Methylumbelliferyl butyrate (4-MUB) fluorescent assay detecting tributyrin hydrolysis activity of LesB, LesBMUT2 and EV (empty vector) proteins expressed in E. coli.\n\nUsing 4-MUB assay we were able to detect tributyrin hydrolysis activity and confirm what we observed in tributyrin agar plate. Crude protein extracted from LesBMUT2 expressing E. coli showed a significant increase in tributyrin hydrolysis activity as depicted in Figure 3b.\n\nAsn228 in LipA is an interesting residue in the context of tributyrin hydrolysis. This residue is involved in the binding of glycoside detergent β-octyl glucoside (BOG) in LipA (PDBid:3H2K) in X. oryzae pv oryzae. Interestingly, while the mutant N228W mutant (PDBid:3H2I) is still able to hydrolyze tributyrin, it showed virulence deficiency similar to that of the LipA-deficient strain BXO200136. Corroborating the redundancy of Asn228 in tributyrin hydrolysis, we note that this residue falls outside the 8 Å radius chosen for LesBMUT2 mutations, which were able to induce tributyrin hydrolysis.\n\n\nMaterials and methods\n\nIn order to test for tributyrin hydrolysis activity, the new open reading frame with suggested mutations as well as wild type LesB was codon optimized for expression in E. coli and were chemically synthesized (DNA2.0, Menlo Park, CA). The resulting coding sequences were cloned into pJ401: T5 expression vector from DNA2.0 to obtain high amounts of recombinant LesB4 and LesB.\n\nE. coli strains expressing LesB wild type and LesB4 as well as empty vector control (EV) were inoculated into liquid cultures and grown overnight at 37°C with constant shaking at 200 rpm. The overnight cultures were added to a larger flask with fresh LB media and grown under the same conditions until they reached OD of 0.5–0.8, at the point which 0.3 mM of IPTG (Sigma Aldrich) was added to each culture to induce the promoter. Induced cultures were incubated at room temperature with constant shaking at 200 RPM overnight. The following day cultures were spun down at 5000g (Sigma 3K10) to pellet the cells and resulting pellet was suspended in sterile PBS. Next, the obtained cells in PBS were lysed using a microfluidizer (Microfluidics M-110L) machine. The extracted protein was quantified and used for the 4-MUB assay.\n\nThe quantitative detection of LesA enzyme was carried out using MUB Assay based on Vaneechoutte et al.37. For each samples 3 technical replicates were used. Briefly, A stock solution of the substrate, made by dissolving 10 mg of 4-methylumbelliferyl butyrate (4-MUB) (Sigma Chemical Co., St. Louis, MO) in 1 mL of dimethyl sulfoxide (DMSO) and 10 μL of Triton X-100 to obtain 40 mM stock solution which was further diluted to 5 mM using 0.1 M citrate buffer (pH 5.0). For each reaction 80 μL of substrate was added to 20 μL of each sample immediately before the fluorescent intensity was read in a fluorometer Plate Reader SpectraMax M2 (Molecular Devices) at 365 nm excitation and 455 nm emission at 30°C for 30 minutes. Fluorescence values measured in 4-MUB assay were used to calculate mean and standard deviations. After subtracting the background (PBS) values were plotted using Prism version 6.0c.\n\nIn silico methods. The current work provides in vivo validation of previously described in silico methods26,31,32. The underlying theoretical foundation for our methods is the non-triviality of the spatial and electrostatic congruence in cognate pairs seen across various structures of the same catalytic function. We identify spatially equivalent residues that have differing electrostatic properties, based on the logic that functional divergence in the protein family arises from these residues.\n\nIn order to superimpose the two scaffolds - LesA=(Asp336, Ser176 and His377) and LesB=(Asp323, Ser165 and His365) - we applied linear and rotational transformations for all atoms in LesA and LesB such that the Cβ atoms of all residues lay on the same plane (Z=0), Asp336LesACβ and Asp323LesBCβ were at the center, and Ser176LesACβ and Ser165LesBCβ lie on the Y axis. This superimposition was outputted as a Pymol formatted file.\n\nWe aligned the residues from LesA which are within 7 Å radius from the active site residues (Asp336, Ser176 and His377). The choice of the radial distance that encompasses interacting residues is critical, since a small radius will not include enough residues, and a large one will include irrelevant ones. Next, for LesB we identified residues that were in the vicinity of each of the residues in LesA, choosing the closest residue as the alignment for the position p. This is possible since we have a consolidated spatial reference frame for both the proteins. RaptorX was then used to predict the structure of the mutated LesB33.\n\nAdaptive Poisson-Boltzmann Solver (APBS) and PDB2PQR packages were used to calculate the potential difference between the reactive atoms of the corresponding proteins38,39. The APBS parameters and electrostatic potential units were set as described previously in26.\n\nAll protein structures were rendered by PyMol (http://www.pymol.org/). The alignment and cladograms images were created using Seaview40. PHYML was used to generate phylogenetic trees from these alignments, which searches for a tree with the highest probability or likelihood that, given a proposed model of evolution and the hypothesized history, would give rise to the observed data set (method of maximum likelihood)41.\n\n\nData availability\n\nF1000Research: Dataset 1. LipA and LesA/B/C multiple sequence alignment and 4-MUB assay data, 10.5256/f1000research.5147.d3495442", "appendix": "Author contributions\n\n\n\nHG performed the experiments. SC wrote the computer programs. All authors analyzed the data, and contributed equally to the writing and subsequent refinement of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAMD wishes to acknowledge grant support from the California Department of Food and Agriculture PD/GWSS Board. BJ acknowledges financial support from Tata Institute of Fundamental Research (Department of Atomic Energy). Additionally, BJR is thankful to the Department of Science and Technology for the JC Bose Award Grant. BA acknowledges financial support from the Science Institute of the University of Iceland.\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\nConant GC, Wolfe KH: Turning a hobby into a job: how duplicated genes find new functions. Nat Rev Genet. 2008; 9(12): 938–950. PubMed Abstract \| Publisher Full Text\n\nLynch M, Conery JS: The evolutionary fate and consequences of duplicate genes. Science. 2000; 290(5494): 1151–1155. PubMed Abstract \| Publisher Full Text\n\nZhang J: Evolution by gene duplication: an update. Trends Ecol Evol. 2003; 18(6): 292–298. Publisher Full Text\n\nBergthorsson U, Andersson DI, Roth JR: Ohno’s dilemma: evolution of new genes under continuous selection. Proc Natl Acad Sci U S A. 2007; 104(43): 17004–17009. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNasvall J, Sun L, Roth JR, et al.: Real-time evolution of new genes by innovation, amplification, and divergence. Science. 2012; 338(6105): 384–387. PubMed Abstract \| Publisher Full Text\n\nMoran GP, Coleman DC, Sullivan DJ: Comparative genomics and the evolution of pathogenicity in human pathogenic fungi. Eukaryotic Cell. 2011; 10(1): 34–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHube B, Stehr F, Bossenz M, et al.: Secreted lipases of Candida albicans: cloning, characterisation and expression analysis of a new gene family with at least ten members. Arch Microbiol. 2000; 174(5): 362–374. PubMed Abstract \| Publisher Full Text\n\nGacser A, Trofa D, Schafer W, et al.: Targeted gene deletion in Candida parapsilosis demonstrates the role of secreted lipase in virulence. J Clin Invest. 2007; 117(10): 3049–3058. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoigt CA, Schafer W, Salomon S: A secreted lipase of Fusarium graminearum is a virulence factor required for infection of cereals. Plant J. 2005; 42(3): 364–375. PubMed Abstract \| Publisher Full Text\n\nAparna G, Chatterjee A, Sonti RV, et al.: A cell wall-degrading esterase of Xanthomonas oryzae requires a unique substrate recognition module for pathogenesis on rice. Plant Cell. 2009; 21(6): 1860–1873. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimpson AJ, Reinach FC, Arruda P, et al.: The genome sequence of the plant pathogen Xylella fastidiosa. The Xylella fastidiosa Consortium of the Organization for Nucleotide Sequencing and Analysis. Nature. 2000; 406(6792): 151–159. PubMed Abstract \| Publisher Full Text\n\nAlston JM, Fuller KB, Kaplan JD, et al.: Assessing the returns to R&D on perennial crops: the costs and benefits of pierce’s disease research in the california winegrape industry. Aust J Agric Resour Econ. 2014. Publisher Full Text\n\nStrange RN, Scott PR: Plant disease: a threat to global food security. Annu Rev Phytopathol. 2005; 43: 83–116. PubMed Abstract \| Publisher Full Text\n\nHopkins D, Purcell A: Xylella fastidiosa: cause of Pierce’s disease of grapevine and other emergent diseases. Plant Disease. 2002; 86(10): 1056–1066. Publisher Full Text\n\nGray D, Li Z, Hopkins D, et al.: Transgenic grapevines resistant to Pierce’s disease. HortScience. 2005; 40(4): 1104–1105. Reference Source\n\nAgueero CB, Uratsu SL, Greve C, et al.: Evaluation of tolerance to Pierce’s disease and botrytis in transgenic plants of Vitis vinifera l. expressing the pear pgip gene. Mol Plant Pathol. 2005; 6(1): 43–51. PubMed Abstract \| Publisher Full Text\n\nDandekar AM, Gouran H, Ibanez AM, et al.: An engineered innate immune defense protects grapevines from Pierce disease. Proc Natl Acad Sci U S A. 2012; 109(10): 3721–3725. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolon DN, Mayo SL: Enzyme-like proteins by computational design. Proc Natl Acad Sci U S A. 2001; 98(25): 14274–14279. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJiang L, Althoff EA, Clemente FR, et al.: De novo computational design of retro-aldol enzymes. Science. 2008; 319(5868): 1387–1391. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaiella M, Andreozzi C, de Rosales RT, et al.: An artificial di-iron oxo-protein with phenol oxidase activity. Nat Chem Biol. 2009; 5(12): 882–884. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSiegel JB, Zanghellini A, Lovick HM, et al.: Computational design of an enzyme catalyst for a stereoselective bimolecular Diels-Alder reaction. Science. 2010; 329(5989): 309–313. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRothlisberger D, Khersonsky O, Wollacott AM, et al.: Kemp elimination catalysts by computational enzyme design. Nature. 2008; 453(7192): 190–195. PubMed Abstract \| Publisher Full Text\n\nReetz MT, Carballeira JD: Iterative saturation mutagenesis (ISM) for rapid directed evolution of functional enzymes. Nat Protoc. 2007; 2(4): 891–903. PubMed Abstract \| Publisher Full Text\n\nClimie S, Ruiz-Perez L, Gonzalez-Pacanowska D, et al.: Saturation site-directed mutagenesis of thymidylate synthase. J Biol Chem. 1990; 265(31): 18776–18779. PubMed Abstract\n\nReetz MT, Carballeira JD, Peyralans J, et al.: Expanding the substrate scope of enzymes: combining mutations obtained by CASTing. Chemistry. 2006; 12(23): 6031–6038. PubMed Abstract \| Publisher Full Text\n\nChakraborty S, Minda R, Salaye L, et al.: Active site detection by spatial conformity and electrostatic analysis--unravelling a proteolytic function in shrimp alkaline phosphatase. PLoS One. 2011; 6(12): e28470. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Ásgeirsson B, Minda R, et al.: Inhibition of a cold-active alkaline phosphatase by imipenem revealed by in silico modeling of metallo-β-lactamase active sites. FEBS Lett. 2012; 586(20): 3710–3715. PubMed Abstract \| Publisher Full Text\n\nRendon-Ramirez A, Shukla M, Oda M, et al.: A computational module assembled from different protease family motifs identifies PI PLC from Bacillus cereus as a putative prolyl peptidase with a serine protease scaffold. PLoS One. 2013; 8(8): e70923. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rendon-Ramirez A, Ásgeirsson B, et al.: Dipeptidyl peptidase-IV inhibitors used in type-2 diabetes inhibit a phospholipase c: a case of promiscuous scaffolds in proteins [v1; ref status: approved 1, approved with reservations 1, http://f1000r.es/2hw]. F1000Research. 2013; 2: 286. Reference Source\n\nChakraborty S, Rao BJ: A measure of the promiscuity of proteins and characteristics of residues in the vicinity of the catalytic site that regulate promiscuity. PLoS One. 2012; 7(2): e32011. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S: An automated flow for directed evolution based on detection of promiscuous scaffolds using spatial and electrostatic properties of catalytic residues. PLoS One. 2012; 7(7): e40408. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rao BJ, Baker N, et al.: Structural phylogeny by profile extraction and multiple superimposition using electrostatic congruence as a discriminator. Intrinsically Disordered Proteins. 2013; 1: e25463. Publisher Full Text\n\nPeng J, Xu J: RaptorX: exploiting structure information for protein alignment by statistical inference. Proteins. 2011; 79(Suppl 10): 161–171. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKonagurthu AS, Whisstock JC, Stuckey PJ, et al.: MUSTANG: a multiple structural alignment algorithm. Proteins. 2006; 64(3): 559–574. PubMed Abstract \| Publisher Full Text\n\nChakraborty S, Rao BJ, Ásgeirsson B, et al.: The electrostatic profile of consecutive Cβ atoms applied to protein structure quality assessment [v2; ref status: awaiting peer review, http://f1000r.es/2cf]. F1000Research. 2013; 2: 243. Publisher Full Text\n\nRajeshwari R, Jha G, Sonti RV: Role of an in planta-expressed xylanase of Xanthomonas oryzae pv. oryzae in promoting virulence on rice. Mol Plant Microbe Interact. 2005; 18(8): 830–837. PubMed Abstract \| Publisher Full Text\n\nVaneechoutte M, Verschraegen G, Claeys G, et al.: Rapid identification of Branhamella catarrhalis with 4-methylumbelliferyl butyrate. J Clin Microbiol. 1988; 26(6): 1227–1228. PubMed Abstract \| Free Full Text\n\nBaker NA, Sept D, Joseph S, et al.: Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98(18): 10037–10041. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDolinsky TJ, Nielsen JE, McCammon JA, et al.: PDB2PQR: an automated pipeline for the setup of Poisson-Boltzmann electrostatics calculations. Nucleic Acids Res. 2004; 32(Web Server issue): W665–667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGouy M, Guindon S, Gascuel O: SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Mol Biol Evol. 2010; 27(2): 221–224. PubMed Abstract \| Publisher Full Text\n\nGuindon S, Lethiec F, Duroux P, et al.: PHYML Online--a web server for fast maximum likelihood-based phylogenetic inference. Nucleic Acids Res. 2005; 33(Web Server issue): W557–559. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGouran H, Chakraborty S, Rao BJ, et al.: LipA and LesA/B/C multiple sequence alignment and 4-MUB assay data. F1000Research. 2014. Data Source" }	[ { "id": "6084", "date": "24 Sep 2014", "name": "Rajan Sankaranarayanan", "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 study, the authors have modeled the duplicated secreted virulence factors of Xylella fastidiosa (LesA, LesB, LesC), based on the closely related LipA protein structure from Xanthomonas oryzae. LesA protein can hydrolyze tributyrin while LesB and LesC cannot, although LesB and LesC both have a catalytic triad (Asp, Ser, His) similar to LesA and LipA. The authors have used a previously developed computational method by their group that suggests mutations based on the spatial and electrostatic properties of the active site residues. In this study, LesA active site was used to suggest mutations in LesB. After the introduction of these mutations, LesB shows activity upon tributyrin. This piece of work is a good example theoretically predicting functionality that can narrow down the search in the sequence space. Comments/ Suggestions:Were tributyrin hydrolysis assays done for LipA and LesA? A comparison would portray the differential effect of mutations. In figure 3(a), the third image was mislabeled as LesBMUT1 instead of LesBMUT2.What does DECAAF stands for? How the spatially equivalent residues were identified which differ in their electrostatic properties? A brief explanation about how this algorithm works would have been better to appreciate the work.The title needs to be changed as it conceptually clashes with 'Directed evolution' used in common parlance, which is experiment based but a black box methodology. Here, they use certain available structural information and hence should not be confused with the above term.", "responses": [] }, { "id": "7453", "date": "16 Feb 2015", "name": "Tomomichi Fujita", "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, the authors clearly demonstrate usefulness of their in silico methods to identify residues which are spatially equivalent but electrostatically different in cognate pairs in protein family, then provide in vivo validation of the method by measuring the enzymatic activity of the designed protein. Such “focused mutation” approach and the experimental validation allow us to identify minimal mutations, which are important for the enzymatic activity and also may shed light on the fate of, for example, the duplicated gene products; pseudogenization, subfunctionalization or neofunctionalization.Minor comments for improvement of article.The authors aimed to induce tributyrin hydrolysis in LesB using minimal mutations. They described either three (MUT1) or eight mutations (MUT2) in LesB protein, where MUT1 did not show tributyrin hydrolysis activity, but MUT2 did. However, I wonder if the number of mutations, eight could be minimal mutations or not. If the authors have other mutations introduced, such as between four and seven mutations in LesB, please explain them. In Fig. 1(a), amino acid positions are numbered, however, to which protein are the numbers? Also, please mark three active site residues of LipA (serine catalytic triad) in the figure. In Fig. 2(c), please indicate five amino acid residues (G363, E364, F366, F367 and T368) in another contiguous stretch. Also, please show the catalytic triad as well. In Table 1, please indicate to which amino acid residues “a, b, c, d” of active site atoms correspond, respectively. In Table 1, I don’t understand amino acid positions, SER164, HIS364, and SER74 in LesBMUT1, which are different from the positions, SER165, HIS365, and ARG/SER75 in LesB/LesBMUT1, respectively. In text, page 5, a right column, line 18 from bottom, “…(LesBMUT1/LesBMUT2) modulate this EPD back to moderately positive again (+12.4 EPD units).” would be “…(LesBMUT1/LesBMUT2) modulate this EPD back to moderately positive again (+19.3 and +12.4 EPD units, respectively).” In text, page 6, a left column, it would be better to show the results of LesBMUT1, showing no hydrolysis activity in Fig. 3(a), (b). If the authors have the data of LesA, please show them together in Fig. 3(a), (b). In Fig. 3(a), LesBMUT1 should be LesBMUT2. In Fig. 3(b) and in Materials and methods, please explain what LesB4 is. In Table 3, whereas the EPD in the pair ‘ad’ of LESA and LESBMUT1 are negative, the EPD in the one of LESBMUT2 is positive (+5.4), showing inconsistency of the presence of the enzymatic activity in LESA and in LESBMUT2, but not in LESBMUT1. Please explain this. In Materials and methods, cell culture and protein extraction, the extracted protein was quantified and used for the 4-MUB assay. Please describe how the authors normalized protein concentration in each sample to calculate mean and standard deviation values for the 4-MUB assay. In Materials and methods, a right column, line 14, what does ‘the position p’ mean, which is described as in “the alignment for the position p”?", "responses": [] } ]	1	https://f1000research.com/articles/3-215
https://f1000research.com/articles/3-213/v1	08 Sep 14	{ "type": "Research Note", "title": "Chief’s seminar: turning interns into clinicians", "authors": [ "Christopher Dittus", "Vanya Grover", "Georgia Panagopoulos", "Kenar Jhaveri", "Vanya Grover", "Georgia Panagopoulos", "Kenar Jhaveri" ], "abstract": "Background: Recent changes in healthcare delivery have necessitated residency education reform. To adapt to these changes, graduate medical education can adopt a chief resident-led clinical curriculum. Chief residents are ideal clinical instructors, as they are recent graduates who have excelled in their residency programs. To effectively use the limited time available for education, chief residents can implement active learning techniques. We present a chief resident-led, small-group, problem-based curriculum for teaching first-year internal medicine residents, and provide preliminary data supporting the efficacy of this approach.Methods: The seminar consisted of 11 4-week modules. Week 1 was a team-based crossword competition. Weeks 2-4 were small-group, problem-based clinical reasoning sessions taught by chief residents. The program was evaluated via pre- and post-module multiple-choice tests. Resident satisfaction data were collected via self-reported, anonymous surveys.Results: Preliminary results revealed a statistically significant increase from pre-test to post-test score for 9 of the 11 modules. The chest pain, fever, abdominal pain, shock, syncope, jaundice, dizziness, anemia, and acute kidney injury modules achieved statistical significance. Additionally, resident satisfaction surveys show that this teaching approach was an enjoyable experience for our residents.Discussion: Our chief seminar is an evidence-based, clinical reasoning approach for graduate medical education that uses active learning techniques. This is an effective and enjoyable method for educating internal medicine residents. Because of its reproducibility, it can be applied throughout residency education.", "keywords": [ "The changing landscape of healthcare delivery has made residency education reform a necessity1. In the hospital", "residents must navigate an increasingly complicated healthcare system", "while having to expeditiously diagnose", "treat", "and discharge patients. Furthermore", "medical knowledge is increasing in both the basic and clinical sciences. This is evidenced by the dramatic rise in the quantity of publications in the medical literature", "particularly randomized controlled trials2. To keep pace with these demands", "graduate medical education can adopt a chief resident-led clinical curriculum. Chief residents are ideal clinical instructors", "as they have the knowledge base of an attending while still understanding the learning needs of residents. To effectively use the limited time available for education", "chief residents can implement active learning techniques. Approaches that fall under active learning are: problem-based learning3–6", "collaborative group work", "and peer instruction7", "8. Small group learning is an effective way to apply these techniques9", "10. The application of active learning principles to medical education has been increasingly promoted in the medical literature1", "11–13. Active learning also allows for the addition of creative modalities", "such as concept mapping", "games", "puzzles", "and extrinsic rewards4", "9", "14", "15. We present a chief resident-led", "small-group", "problem-based curriculum for teaching first-year internal medicine residents", "and provide preliminary data supporting the efficacy of this approach." ], "content": "Background\n\nThe changing landscape of healthcare delivery has made residency education reform a necessity1. In the hospital, residents must navigate an increasingly complicated healthcare system, while having to expeditiously diagnose, treat, and discharge patients. Furthermore, medical knowledge is increasing in both the basic and clinical sciences. This is evidenced by the dramatic rise in the quantity of publications in the medical literature, particularly randomized controlled trials2. To keep pace with these demands, graduate medical education can adopt a chief resident-led clinical curriculum. Chief residents are ideal clinical instructors, as they have the knowledge base of an attending while still understanding the learning needs of residents. To effectively use the limited time available for education, chief residents can implement active learning techniques. Approaches that fall under active learning are: problem-based learning3–6; collaborative group work; and peer instruction7,8. Small group learning is an effective way to apply these techniques9,10. The application of active learning principles to medical education has been increasingly promoted in the medical literature1,11–13. Active learning also allows for the addition of creative modalities, such as concept mapping, games, puzzles, and extrinsic rewards4,9,14,15. We present a chief resident-led, small-group, problem-based curriculum for teaching first-year internal medicine residents, and provide preliminary data supporting the efficacy of this approach.\n\n\nProgram description\n\nParticipants included the intern (i.e. postgraduate year 1, PGY-1) class at a large, urban, tertiary-care hospital. The PGY-1 class consisted of 16 preliminary interns and 27 categorical interns, for a total of 43 participants. During any given module, 16 to 18 interns on the general medical floor participated in the teaching sessions and program evaluation. Week one of the seminar included interns, as well as PGY-2 and PGY-3 residents.\n\nThe chief’s seminar curriculum consisted of eleven, four-week long modules. The module topics were: dyspnea, chest pain, fever, abdominal pain, shock, altered mental status (AMS), syncope, jaundice, dizziness, anemia, and acute kidney injury (AKI). Each four-week long module consisted of four, one-hour weekly sessions. The chest pain and AMS modules were exceptions, having had only three sessions because of scheduling conflicts. Every module began with a competitive session during Week 1. Weeks 2 through 4 were composed of small-group discussions examining module subtopics in greater depth.\n\nWeek 1: The first week of the seminar began with a crossword competition that served as an important tool for engaging and motivating residents. All general medical floor interns and residents were included in this competition. The participating house-staff were divided into approximately 6 groups of 4. Each member of the team was given a module-specific crossword puzzle covering all aspects of the chief complaint (Figure 1). The crossword puzzle was created by the chief residents using a free, downloadable program (http://www.eclipsecrossword.com). Each month a new crossword puzzle was created, with each puzzle taking 1 to 2 hours to prepare. During this hour-long session, team members worked cooperatively while competing with other teams to complete the crossword puzzle. The team that completed the greatest number of crossword puzzle questions, in the shortest amount of time, won the crossword competition for that module. At the end of the session, all of the answers were reviewed with the house-staff. In addition to the motivation garnered inherently by the competition, teams also competed for $40 worth of gift cards. A photograph of the winning team was distributed via email to the Department of Medicine.\n\nThis is an example of a crossword puzzle from the chief seminar module on dizziness16. The questions were targeting high-yield board review topics. Crossword-generating software is readily available via many different websites. We used a free, downloadable program from the following website: http://www.eclipsecrossword.com.\n\nWeeks 2–4: The content of the subtopic weeks varied depending on the module, but the same techniques were employed. The emphasis during these sessions was on the generation of a differential diagnosis based on a clinical reasoning algorithm. By using a problem-based approach (i.e. dizziness, not vertigo), the course created a real-life clinical scenario. Prior to the session, a practical clinical reasoning algorithm was created by the chief residents, with each session taking 2 to 3 hours to prepare. At the start of the session, the interns were separated into two, chief resident-led groups. Each subtopic session (three per module) was initiated by drawing the full concept map (Figure 2). The concept map began with the module title (e.g. dizziness), and then a broad differential diagnosis was determined using an evidence-based clinical algorithm. The algorithm used information gathered from multiple sources, including the history, physical exam, laboratory data, and imaging. For the dizziness module, the initial differential diagnosis was narrowed by using historical questions, then physical exam findings, and lastly by asking targeted questions. Once the concept map was fully developed, a more detailed discussion of one diagnostic pathway (e.g. vertigo) commenced. For each subtopic session, a different diagnostic pathway was discussed in detail. By the end of the module, the intern had an understanding of each diagnosis, as well as an understanding of how the final, targeted diagnosis was derived from the initial, broad differential. The concept map could be handwritten or created via an internet-based program (https://bubbl.us/).\n\nThis is an example of a completed concept map from the chief seminar module on dizziness16. The concept map starts at the top and progresses downward as more information is collected via intern participation and chief resident guidance. Many online concept mapping programs are available. For this seminar, we used the following website: https://bubbl.us/.\n\n\nMethods\n\nTo evaluate the efficacy of our didactic curriculum, we collected pre- and post-course five-question multiple choice tests for each of our modules. Informed consent forms were distributed to all participants and IRB approval for exemption was obtained from Lenox Hill Hospital, North-Shore LIJ (IRB#: 13-045A). Prior to beginning a module, each intern received a unique identifier. This number was used to link the pre- and post-test for each intern participating in the module. The pre-test was given prior to the week 1 crossword competition. The post-test was given after the week 4 teaching session. On completion of a module, pre- and post-test data were entered into a secure, anonymous database according to each unique identifier. Using SPSS Version 20 (IBM SPSS, Chicago IL), data for each module were analyzed via the Wilcoxon Signed-Rank Test. The pre-test and post-test were identical, in order to control for inter-test variability.\n\nA six-item resident satisfaction survey was also distributed for each module (Figure 3). This survey was completely anonymous, and the data collected were descriptive in nature. Survey questions focused on resident satisfaction with the content and style of the module, as well as the perceived effectiveness of the crossword puzzle and small group sessions.\n\nThis is the satisfaction survey evaluation form we used for the dizziness module. We evaluated 6 parameters that were aimed at gaining an overall impression of intern satisfaction with each module.\n\n\nResults\n\nEfficacy results were obtained for all eleven modules (Figure 4). Results showed a statistically significant increase from pre-test to post-test score for 9 of the 11 completed modules. Chest pain, fever, abdominal pain, shock, syncope, jaundice, dizziness, anemia, and AKI achieved statistical significance, while the first module, dyspnea, had a trend towards statistical significance. Additionally, the AMS module had, to a lesser degree, a trend towards significance.\n\nThis figure evaluates the change in score from pre-test to post-test for each of the chief seminar modules. The x-axis lists each module with the number of participants who took both the pre- and post-test. The y-axis has the average participant score (0 to 5). Each module was evaluated for a significant change between pre- and post-test and the p-value can be found at the top of each module’s bar chart.\n\nIntern satisfaction results were obtained for all eleven modules (Figure 5). Survey results were aggregated by survey parameter for each module. Each response was given a numeric code: Strongly Agree (2 points), Agree (1 point), Neither (0 points), Disagree (-1 point), and Strongly Disagree (-2 points). The aggregate results were then weighted according to the numeric code for each response, and averaged. Of the six survey parameters, five (content, style, chief resident effectiveness, improved ability to diagnose, and improved ability to treat) were at or above the “agree” response for all eleven modules. Only the survey question concerning the crossword puzzle had any responses below “agree”.\n\nThis cluster chart displays the satisfaction survey results for each satisfaction parameter by module. The x-axis lists each survey question and the y-axis lists the level of resident satisfaction, with “2” representing strong satisfaction and “-2” representing strong dissatisfaction.\n\n\nDiscussion\n\nThe primary goal of our study was to present a reproducible, chief resident-led, teaching curriculum that applies active learning principles to a clinical reasoning seminar for interns. The interns who participated in our study were focused and engaged during our sessions and their positive satisfaction survey results reflect this. We have described our curriculum in detail with the hope that it can be replicated by other residency programs.\n\nThe second goal of our study was to present preliminary data evaluating the efficacy of this teaching modality. We have shown a statistically significant increase in test scores for 9 of our 11 modules. For these 9 modules, the interns showed retention of clinical reasoning techniques. Coupled with positive satisfaction surveys, we conclude that this curriculum is both effective and desirable. The remaining two modules, dyspnea and AMS, trended towards, but did not achieve, statistical significance. Dyspnea was our first module, and we were not surprised with the lack of statistical significance. Less clear was why AMS did not reach statistical significance. The broad nature of this topic, combined with one less teaching session, likely contributed to the decreased efficacy outcome for this module. It is also important to note that the number of participants in any of these sessions was low, and if there were more participants, statistical significance would have likely been achieved.\n\nSeveral limitations of our study have been identified. First, the lack of a comparison group does not allow us to conclude that our teaching curriculum was more effective than a traditional teaching approach. Being a pilot study, the first step was to show that this was an effective teaching method. Subsequent research can build on our preliminary findings by directly comparing this novel curriculum to a traditional, purely lecture-based curriculum. Another limitation was the potential for a practice effect, where the answers to the pre-test are remembered for the post-test. This is most pronounced with a short interval between pre- and post-testing. We did not review any pre-test answers with the participants, and our tests were separated by one month, which was likely sufficient to limit this bias. Another possibility is that the interns searched for the answers to the pre-test before they completed the post-test. We find this unlikely since the post-test scores never approached 100% accuracy. Lastly, our study did not examine long-term retention. This requires the post-test to be repeated at a longer interval, which will be an area of future research.\n\nDespite these, relatively minor, limitations, we have shown that this chief resident-led, evidence-based clinical reasoning approach to graduate medical education is effective and enjoyable for internal medicine residents and we feel it should be applied and adapted throughout residency education.\n\n\nData availability\n\nfigshare: Chief seminar datasets for pre-/post-testing and satisfaction surveys. Doi: 10.6084/m9.figshare.115769917", "appendix": "Author contributions\n\n\n\nC.D. and V.G.: responsible for the program conception, design, and execution. Responsible for data collection and storage. Primary authors of manuscript responsible for writing, editing and figure creation.\n\nG.P.: responsible for all statistical analysis and interpretation. Edited manuscript.\n\nK.J.: responsible for the program conception and design. Responsible for writing and editing 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 authors would like to thank the 2012–2013 intern class at our hospital for participating in this teaching curriculum, as well as all the residents that participated in our crossword competitions. Additionally, we would like to thank our chairman, Dr. Jack Ansell, for giving us the opportunity to implement a creative approach to medical education.\n\nWe presented an earlier version of this manuscript as a poster at the Association of Program Directors in Internal Medicine (APDIM) annual meeting in Orlando, Florida, in 2013, and as an oral presentation at the Northeastern Group on Educational Affairs (NEGEA) annual retreat in New York City, in 2013.\n\nSmall monetary rewards for the participants came from the internal medicine department at our hospital.\n\n\nReferences\n\nLee E, Lazarus ME, El-Farra N: An updated focus on internal medicine resident education. Am J Med. 2012; 125(11): 1140–3. PubMed Abstract \| Publisher Full Text\n\nDruss BG, Marcus SC: Growth and decentralization of the medical literature: implications for evidence-based medicine. J Med Libr Assoc. 2005; 93(4): 499–501. PubMed Abstract \| Free Full Text\n\nHmelo-Silver CE: Problem-based learning: What and how do students learn? Educ Psychol Rev. 2004; 16(3): 235–66. Publisher Full Text\n\nKinkade S: A snapshot of the status of problem-based learning in U.S. medical schools, 2003–04. Acad Med. 2005; 80(3): 300–1. PubMed Abstract\n\nNeville AJ: Problem-based learning and medical education forty years on. A review of its effects on knowledge and clinical performance. Med Princ Pract. 2009; 18(1): 1–9. PubMed Abstract \| Publisher Full Text\n\nKoh GC, Khoo HE, Wong ML, et al.: The effects of problem-based learning during medical school on physician competency: A systematic review. CMAJ. 2008; 178(1): 34–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeberschock TB, Ginn TC, Reinhold J, et al.: Change in knowledge and skills of Year 3 undergraduates in evidence-based medicine seminars. Med Educ. 2005; 39(7): 665–71. PubMed Abstract \| Publisher Full Text\n\nMichael J: Where’s the evidence that active learning works? Adv Physiol Educ. 2006; 30(4): 159–67. PubMed Abstract \| Publisher Full Text\n\nCook DA: Modern learning principles. In: Williams FK, Colbert C, Costa ST, et al. editors. A Textbook for Today’s Chief Medical Resident. 20th ed. Alexandria, VA: Association of Program Directors in Internal Medicine; 2012. p. 73–84.\n\nWalton H: Small group methods in medical teaching. Med Educ. 1997; 31(6): 459–64. PubMed Abstract \| Publisher Full Text\n\nKaufman DM: ABC of learning and teaching in medicine, applying educational theory in practice. BMJ. 2003; 326: 213–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFitzgibbons JP, Bordley DR, Berkowitz LR, et al.: Association of Program Directors in Internal Medicine. Redesigning residency education in internal medicine: a position paper from the Association of Program Directors in Internal Medicine. Ann Intern Med. 2006; 144(12): 920–6. PubMed Abstract \| Publisher Full Text\n\nProber CG, Heath C: Lecture halls without lectures – a proposal for medical education. N Engl J Med. 2012; 366(18): 1657–9. PubMed Abstract \| Publisher Full Text\n\nCalderon KR, Vij RS, Mattana J, et al.: Innovative teaching tools in nephrology. Kidney Int. 2011; 79(8): 797–9. PubMed Abstract \| Publisher Full Text\n\nRondon H, Jhaveri KD: Nephrology crossword: Divalents, a journey through the nephron. Kidney Int. 2012; 82(4): 500–1. Publisher Full Text\n\nPost RE, Dickerson LM: Dizziness: a diagnostic approach. Am Fam Physician. 2010; 82(4): 361–8, 369. PubMed Abstract\n\nDittus C, Grover V, Panagopoulos G, et al.: Chief seminar datasets for pre-/post-testing and satisfaction surveys. figshare. 2014. Data Source" }	[ { "id": "6066", "date": "25 Sep 2014", "name": "Janet Wale", "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 informative and easily understood report on an observational study of a pilot educational program for teaching 1st-year interns, by chief residents. It shows a number of positive results.It would be interesting to also get feedback from the chief residents for each module; and information on ‘what they would change’ (if anything), particularly as they develop the materials according to a framework.Some of the limitations of the pilot are clearly identified, which is very useful.Background, line 11: a word is missing – should read ‘..knowledge base of an attending clinician, while…’", "responses": [] }, { "id": "6068", "date": "30 Sep 2014", "name": "Stewart Babbott", "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\nInternship is a time of rapid professional and personal growth. Instruction which meets the practical needs of the learners and delivered in ways concordant with the practice environment can be more likely to viewed as successful. This manuscript describes a year long program (11 modules of 4 weeks each) developed for interns to address common clinical situations and diagnoses.Overall, this was a well thought out and delivered program. The authors sought ways to engage the learners including basing the instruction on active learning techniques.Title: Turning interns into clinicians is the goal, but the program only brings the interns part of the way to that goal. The title may overstate the scope of the program; however, the authors note this was a pilot program.Abstract: Provides an adequate summary.Study design, methods and analysis are appropriate. The goals of the study as stated were to deveop a reproducable chief led program that applied active learning principles to a clinical reasoning seminar (done by survey). A secondary goal was to present preliminary data to evaluate efficacy (done by pre-post scores).Discussion and conclusions. The authors appropriately discuss the findings, limitations and conclusions; the curriculum was well received and post test scores improved for most of the modules.Replication: the authors have supplied their surveys, and example of the crossword, the topics and the format within the residency program to allow for replication of the process of the program.Comment: to follow this study on to further assess the process of turning interns into clinicians, it would be of interest to assess their clinical performance. One approach could be a review of their patients (inpatient and outpatient) for diagnoses addressed and patient outcomes. Another approach is development of an Entrustable Professional Activity around these key areas, and to assess housestaff performance in the relevant venue (eg: shock in the ICU or jaundice in the hospital). I liked the use of the concept map that the learners developed for each module: one way to potentially assess that is to select cases for case conference or morning report, had have the interns use their concept map in the discussion of that case.", "responses": [] } ]	1	https://f1000research.com/articles/3-213
https://f1000research.com/articles/3-211/v1	04 Sep 14	{ "type": "Case Report", "title": "Case Report: Primary Leiomyosarcoma of the breast with unusual metastasis to the femur", "authors": [ "Evgeniya Sokolovskaya", "Zheng Liu", "Kelly Weintraub", "Arpad Szallasi", "Yasmeen Shariff", "Evgeniya Sokolovskaya", "Zheng Liu", "Kelly Weintraub", "Yasmeen Shariff" ], "abstract": "With less than 40 cases reported, primary leiomyosarcoma is an extremely rare form of breast cancer (less than 0.0006% of cases) with unpredictable biological behavior that usually presents as a slow growing, mobile mass in middle age women. Most cases are low-grade and are cured by complete excision with wide margins. After surgical resection, late local recurrence and distant hematogenous metastasis to lungs and liver is, however, well-documented. To the best of our knowledge, bone metastasis has never been reported. Here we present a case of primary leiomyosarcoma of the breast metastatic to the femur.", "keywords": [ "A 58 year-old woman (G4P2) with no prior mammograms presented with complaint of increasing pain in her right breast for 7 months. Physical examination revealed an enlarged breast with multiple visible nodules but no adenopathy. Mammography detected a large mass associated with calcifications and thickening of the overlying skin (BIRADS 5) (Figure 1). The left breast was normal. Sonographically", "the mass was primarily hypoechoic (Figure 2). MRI with contrast showed a lobulated", "heterogeneously enhancing mass involving most of the right breast with multiple areas of necrosis. No lymphadenopathy or chest wall involvement was seen (Figure 3)." ], "content": "Case report\n\nA 58 year-old woman (G4P2) with no prior mammograms presented with complaint of increasing pain in her right breast for 7 months. Physical examination revealed an enlarged breast with multiple visible nodules but no adenopathy. Mammography detected a large mass associated with calcifications and thickening of the overlying skin (BIRADS 5) (Figure 1). The left breast was normal. Sonographically, the mass was primarily hypoechoic (Figure 2). MRI with contrast showed a lobulated, heterogeneously enhancing mass involving most of the right breast with multiple areas of necrosis. No lymphadenopathy or chest wall involvement was seen (Figure 3).\n\nThere are scattered calcifications within the mass and overlying skin thickening. This corresponds to the visible and palpable abnormality and is strongly suspicious of malignancy (BIRADS 5).\n\nIn the right breast, there is a large lobulated heterogeneously enchancing mass with multiple areas of non-enchancement consistent with extensive necrosis. On axial image, the mass measures 15 cm × 9 cm × 13 cm. In the left breast, there is no suspicious mass. No suspicious adenopathy is seen in the axillae bilaterally.\n\nUltrasound guided core biopsy of the right breast revealed a spindle-cell neoplasm composed of tumor cells with blunt ended nuclei that were strongly positive for smooth muscle actin (SMA) and lacked expression of pan-cytokeratin, CD34, and S-100 (not shown). This immunophenotype is most consistent with a diagnosis of breast sarcoma. Metastatic workup detected small bilateral lung nodules.\n\nIn 2011 the patient underwent right total mastectomy with partial resection of the pectoralis muscle without chemo- or radiation therapy. Gross examination of the mastectomy specimen revealed a large (15 cm), firm, well-circumscribed mass. Microscopically, the tumor was composed of relatively bland spindle cells arranged as intersecting fascicles. The tumor was positive for SMA and vimentin, and negative for desmin, S-100, CD34, pan-cytokeratin, and neuron-specific enolase. A diagnosis of leiomyosarcoma was made. The resection margins were clean (> 1 cm).\n\nTwo years later, the patient returned with a deep aching pain in her right knee and lower thigh. An X-Ray of her right femur showed a large lucent lesion with endosteal scalloping, suspicious for metastatic disease (Figure 4). A repeat nuclear bone scan was positive for a new increased radiotracer uptake in the right femur. A CT of the chest, abdomen and pelvis discovered a new 3 cm soft tissue mass within the soft tissues in the right gluteal region and multiple lung nodules that were either new or have increased in size compared to previous CTs (Figure 5). An ultrasound guided right gluteal mass full-core biopsy revealed a spindle cell neoplasm similar to the previously excised breast leiomyosarcoma, confirming the diagnosis of metastatic disease (Figure 6). The metastatic gluteal and femoral tumors were resected and chemotherapy with Gemzar (gemcitabine, Eli Lilly) and Taxotere (docetaxel, Sanofi-Aventis) in a 21-day cycle was initiated. Despite chemotherapy, the lung nodules have been increasing in number and size (noted 2 months after the start of the chemotherapy).\n\nA. US guided right gluteal mass full-core biopsy with an 18-gauge BioPince. B. Metastatic neoplastic spindle shaped cells in the right gluteal lesion (H&E, 400X).\n\n\nDiscussion\n\nPrimary leiomyosarcoma of the breast is an extremely rare malignant neoplasm of uncertain biological behavior. There are less than 40 well-documented cases reported in the English medical literature1. The majority of these cases presented as a well-circumscribed mass in the breast of postmenopausal women, although it has also been described in adolescent girls. The histogenesis of the entity is not clear. The myofibroblasts in the nipple areola complex have been proposed as the origin for the neoplasm2. Most reported cases were relatively indolent but aggressive behavior with local recurrence and distant hematogenous metastasis to lungs and liver is also well-documented3. The mainstay treatment is wide margin local excision. Most reported cases have undergone mastectomy with a few exceptions being treated with lumpectomy3. Axillary dissection is believed to be unnecessary as the primary leiomyosaroma of the breast does not spread through the lymphatic route.\n\nWith a size of 15 cm, the present case represents the third largest tumor of all documented cases. Although bone is a common metastatic site for breast carcinoma, to the best of our knowledge breast leiomyosarcoma metastatic to the bone has not been reported. Prognostic factors predicting aggressive biological behavior in mammary leiomysarcomas are yet to be established4,5.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nThe imaging studies were performed by Drs. Sokolovskaya and Shariff. The pathology work-up was done by Drs. Liu and Weintraub. The diagnostic pathology slides were reviewed and the manuscript was written by Dr. Szallasi. All authors were involved in critically revising the manuscript and approved the final version for publication.\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\nNagao T, Hojo T, Tanaka-Akashi S, et al.: Primary leiomyosarcoma of the breast. Breast J. 2012; 8(1): 81–82. PubMed Abstract \| Publisher Full Text\n\nCameron HM, Hamperl H, Warambo W: Leiomyosarcoma of the breast originating from myothelium (myoepithelium). J Pathol. 1974; 114(2): 89–92. PubMed Abstract \| Publisher Full Text\n\nRane SU, Batra C, Saikia UN: Primary leiomyosarcoma of breast in an adolescent girl: a case report and review of the literature. Case Rep Pathol. 2012; 2012: 491984. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdem C, Reynolds C, Ingle JN, et al.: Primary breast sarcoma: clinicopathologic series from the Mayo Clinic and review of the literature. Br J Cancer. 2004; 91(2): 237–241. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPardo-Mindan J, Garcia-Julian G, Eizaguirre Altuna M: Leiomyosarcoma of the breast. Report of a case. Am J Clin Pathol. 1974; 62(4): 477–80. PubMed Abstract" }	[ { "id": "6424", "date": "03 Nov 2014", "name": "Offiong Ikpatt", "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 title, abstract and content are appropriate. No changes are required. For original research, the experimental design, including controls and methods, is adequate; results are presented accurately and the conclusions are justified and supported by the data.", "responses": [] }, { "id": "6425", "date": "17 Nov 2014", "name": "Zsolt Baranyai", "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 title covers the content and the message of the manuscript. The abstract summarizes the content of the article briefly and in a way easy to understand. The abstract is well-constructed and clear. The conclusion is sensible and well-balanced. Although the authors present a case first reported in the literature the conclusion is modest. There are no bias or competing interests.The manuscript meets the criteria for case reports of the journal.It reports the first diagnosed case of bone metastasis of a rare disorder, the leiomyosarcoma of the breast in a well-documented and straightforward way. Reviewing the literature no English article about a case like this was found, thus, the primary rank can be confirmed. The article is valuable not only due to the description of an extremely rare case, but it also calls the attention to the possibility of bone metastasis in this disorder that should be sought for in the future in leiomyosarcoma. I fully support the indexing of the article.", "responses": [] } ]	1	https://f1000research.com/articles/3-211
https://f1000research.com/articles/3-207/v1	02 Sep 14	{ "type": "Case Report", "title": "Case Report: Testicular failure possibly associated with chronic use of methylphenidate", "authors": [ "Ranjith Ramasamy", "Pranav Dadhich", "Ashna Dhingra", "Larry Lipshultz", "Pranav Dadhich", "Ashna Dhingra", "Larry Lipshultz" ], "abstract": "Methylphenidate is a commonly prescribed treatment for attention deficit hyperactivity disorder (ADHD). However, little is known about its adverse effects on the male reproductive system. We report a 20-year-old male patient whose chief complaint was of delayed puberty. He spoke in a high-pitched voice and complained of lack of body hair, impaired libido, inadequate erectile function, chronic fatigue, and low energy. He had been treated with methylphenidate as an infant and had continued treatment for 17 years. On examination, the patient was lean and visibly lacked facial or body hair. He further explained that he had never been able to grow underarm or facial hair and that he was often mistakenly considered a young teenager rather than a 20-year-old. The patient’s genitalia were categorized as Tanner Stage 2. Laboratory studies confirmed low serum follicle-stimulating hormone (FSH), luteinizing hormone (LH), and testosterone levels. The patient was given exogenous testosterone supplementation with pellets and human chorionic gonadotropin to maintain testicular size. After 4 months his symptoms improved and he demonstrated signs of puberty. Our goal is to further elucidate the possible impact of methylphenidate on the male reproductive system.", "keywords": [ "Methylphenidate", "hypogonadism", "delayed puberty" ], "content": "Introduction\n\nThe use of methylphenidate (Ritalin®) has surged over the past decade and is the hallmark therapy for treatment of attention deficit hyperactivity disorder (ADHD)1. Moreover, the number of individuals diagnosed with ADHD has greatly increased within the past decade2. Nevertheless, the long-term effects on various organ systems have not been fully evaluated.\n\nBecause of the increasing incidence of ADHD and the widespread use of methylphenidate in the pediatric population3, it is important to determine whether methylphenidate has any adverse gonadotropic effects. Known adverse effects of methylphenidate include the drug’s impact on the cardiovascular system (high blood pressure, shortness of breath and irregular heart beat), on behavior, and mood (aggression, restlessness, hallucinations, unusual behavior, or motor tics); yet little information exists in the current literature regarding adverse effects of methylphenidate on the human reproductive system. Juvenile Rhesus monkeys treated with high doses of methylphenidate in a controlled experiment displayed a significant delay in testicular descent and were noted to have lower than normal serum testosterone levels4. A parallel experiment was performed using male mice that were given increasing doses of methylphenidate. The mice treated with higher doses of methylphenidate experienced a significant decrease in body weight and reduction in Leydig cell count5, indicating that serum testosterone levels and fertility were significantly reduced.\n\n\nPatient information\n\nThe patient, a 20-year-old Latino male, initially came to the clinic with a chief complaint of “delayed puberty”. In addition, he complained of his high-pitched voice, lack of libido, low energy level, chronic fatigue and poor erectile function. His height was similar to his friends of the same age although he was very lean and could not gain weight. The patient noted that on the basis of his physical appearance people often perceived him to be an adolescent, around the age of 12. He had never had facial or body hair, although he did have some pubic hair. The patient’s past medical history showed use of methylphenidate (dosage varied with age) for approximately 17 years with voluntary cessation a few years ago. The patient currently uses tobacco in the form of an electronic cigar and has been smoking for the past 8 years. He denied alcohol and drug usage. The patient is sexually active and engages in heterosexual sexual activity with no report of sexually transmitted disease. Family history was unremarkable.\n\n\nClinical findings\n\nOn physical examination, the patient had neither underarm hair nor facial hair. He was 180cm and weighed 55kg. The patient’s genitalia were Tanner Stage II, (which is described to be occurring in males aged 8–15 years old and shows enlargement of the testes with pigmentation, minimal to no enlargement of the penis and long, downy hair with a variable pattern). The patient’s bilateral descended testes were 6cc in volume.\n\n\nDiagnostic assessment\n\nHormone testing revealed that the patient had a follicle-stimulating hormone (FSH) value of 3.0 mIU/ml (normal range is 4.0 – 10.0 mIU/ml), a luteinizing hormone (LH) value of 4 mIU/ml (normal range is 6.0 – 19.0 mIU/ml), and a serum total testosterone level of 120 ng/dl (normal values range from 200 to 1000 ng/dl). Semen analysis X 2 showed azoospermia both on initial examination and after use of pellet. The patient has a normal 46XY karyotype and no Y-chromosome microdeletion. The results of laboratory tests were consistent with the patient’s idiopathic testicular failure and warranted further exploration of the link between chronic methylphenidate use and effects on reproductive parameters.\n\n\nTherapeutic intervention\n\nThe patient was advised to begin testosterone supplementation. He chose to receive testosterone pellets (subcutaneous implantable testosterone). Testopel creates a depot of testosterone that is slowly released over a long period of time (~4–6 months). Because exogenous testosterone suppresses intrinsic testosterone synthesis, the patient was given 1500 IU human chorionic gonadotropin (hCG) therapy per week in order to maintain testicular size6.\n\n\nFollow-up and outcomes\n\nThe patient was seen 4 months after his initial visit. He described a desirable increase in energy, increased libido, and better erectile function. Physical examination showed new facial and armpit hair, increased thickness of pubic hair, and maintenance of testicular size. Laboratory testing revealed low FSH (2.0 mIU/ml), and low LH (2 mIU/mL) levels, likely related to exogenous testosterone supplementation. On the other hand, the patient’s serum testosterone increased dramatically to 861 ng/dL. Overall, the patient was very enthusiastic about the progress associated with the therapy and underwent a second insertion of testosterone pellets. As for his fertility, a testicular biopsy for sperm retrieval with assisted reproduction may give him the best chance of fathering a biological child.\n\n\nDiscussion\n\nAlthough methylphenidate has been studied for many years, its effects on male gonadal function have only recently become a topic of interest among clinicians. The patient described in our case study seemed to exhibit characteristics related to the effects of chronic use of methylphenidate on development of human reproductive function.\n\nBecause the developmental changes that occurred in the human subject occurred over a number of years of treatment with methylphenidate, there is very little information about the patient’s condition and how it developed during that period of time. There is also little information for comparison with our observations because there is no much literature available regarding this topic.\n\nOf the few previous studies on the effects of methylphenidate on male gonadal function, one shows results obtained using Wistar rats tested with increasing amounts of methylphenidate. The data show an increase in abnormalities in sperm and a decrease in testicular volume7. Another study using Wistar rats showed that increasing levels of methylphenidate led to increased p53 expression and apoptosis of germ cells8.\n\nThe study involving rats reveals similarities to the case study in question. The case reported describes a possible association between methylphenidate use and testicular failure. The rats were given increasing doses of the drug over a long period of time; our patient was also taking the drug to treat ADHD for a very long time (~14 years). However, while the rats were monitored from the beginning of the drug treatment, the subject in the case study has only recently come under observation after 14 years of methylphenidate use.\n\nAnother of the few existing studies was reported in the Proceedings of the National Academy of Sciences. It was conducted on Rhesus monkeys treated with doses of methylphenidate at designed to mimic the amounts given to human ADHD patients4. This was one of the first recorded studies displaying altered testicular function in primates following methylphenidate treatment. Over a 40 month time span, the monkeys displayed signs of delayed puberty, including impaired testicular descent and smaller-sized testicles, as well as lower testosterone levels9. As noted above, delayed puberty was also observed in our patient, as demonstrated by the Tanner staging and the low testosterone levels in serum. Previous human research yielded similar results. Studies conducted in adolescent boys showed delays in puberty10. The drug of first choice was methylphenidate, although some subjects were given dexamphetamine. Subjects receiving either methylphenidate or dexamphetamine showed lower weight and BMIs coupled with delayed pubertal development10. Although the testicular damage induced by methylphenidate is well recognized, the precise mechanisms underlying its toxicity to the testes remains unclear. It is tempting to speculate that methylphenidate could affect the alpha and beta-adrenergic receptors expressed in the testis6.\n\nThis case report, however, specifically pinpoints methylphenidate and shows an unusual prescription practice for the patient involved. In this case, methylphenidate was started from a very young age compared to normal prescription practices for ADHD. The unknown effects of methylphenidate are currently being studied, but as can be seen, one should exercise caution and patients should be followed closely when prescribing methylphenidate.\n\n\nConsent\n\nWritten informed consent to publish this case report was obtained from the patient.", "appendix": "Author contributions\n\n\n\nRR chose the case and evaluated the patient in the manuscript. PD and AD wrote the draft of the manuscript. LIL supervised the process and critically edited the manuscript. All authors discussed the implications and commented on the manuscript at all stages.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nRR is an NIH K12 Scholar supported by a Male Reproductive Health Research Career (MHRH) Development Physician-Scientist Award (HD073917-01) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Program.\n\n\nReferences\n\nKaranges EA, Stephenson CP, McGregor IS: Longitudinal trends in the dispensing of psychotropic medications in Australia from 2009–2012: Focus on children, adolescents and prescriber specialty. Aust N Z J Psychiatry. 2014. pii: 0004867414538675. PubMed Abstract \| Publisher Full Text\n\nVisser SN, Danielson ML, Bitsko RH, et al.: Trends in the parent-report of health care provider-diagnosed and medicated attention-deficit/hyperactivity disorder: United States, 2003–2011. J Am Acad Child Adolesc Psychiatry. 2014; 53(1): 34–46.e2. PubMed Abstract \| Publisher Full Text\n\nMayes R, Bagwell C, Erkulwater J: ADHD and the rise in stimulant use among children. Harv Rev Psychiatry. 2008; 16(3): 151–66. PubMed Abstract \| Publisher Full Text\n\nMattison DR, Plant TM, Lin HM, et al.: Pubertal delay in male nonhuman primates (Macaca mulatta) treated with methylphenidate. Proc Natl Acad Sci U S A. 2011; 108(39): 16301–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFazelipour S, Jahromy MH, Tootian Z, et al.: The effect of chronic administration of methylphenidate on morphometric parameters of testes and fertility in male mice. J Reprod Infertil. 2012; 13(4): 232–6. PubMed Abstract \| Free Full Text\n\nHsieh TC, Pastuszak AW, Hwang K, et al.: Concomitant intramuscular human chorionic gonadotropin preserves spermatogenesis in men undergoing testosterone replacement therapy. J Urol. 2013; 189(2): 647–50. PubMed Abstract \| Publisher Full Text\n\nMontagnini BG, Silva LS, Dos Santos AH, et al.: Effects of repeated administration of methylphenidate on reproductive parameters in male rats. Physiol Behav. 2014; 133: 122–9. PubMed Abstract \| Publisher Full Text\n\nCansu A, Ekinci O, Ekinci O, et al.: Methylphenidate has dose-dependent negative effects on rat spermatogenesis: decreased round spermatids and testicular weight and increased p53 expression and apoptosis. Hum Exp Toxicol. 2011; 30(10): 1592–600. PubMed Abstract \| Publisher Full Text\n\nMartelle SE, Porrino LJ, Nader MA: Effects of chronic methylphenidate in adolescence on later methylphenidate self-administration in rhesus monkeys. Behav Pharmacol. 2013; 24(5–6): 478–81. PubMed Abstract \| Publisher Full Text\n\nPoulton AS, Melzer E, Tait PR, et al.: Growth and pubertal development of adolescent boys on stimulant medication for attention deficit hyperactivity disorder. Med J Aust. 2013; 198(1): 29–32. PubMed Abstract \| Publisher Full Text" }	[ { "id": "6010", "date": "03 Sep 2014", "name": "Wayland Hsiao", "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\nVery interesting case report of a young man with delayed puberty and primary testicular failure, with no identifiable known cause. The only thing suspicious on their history was long-term use of methylphenidate, with no identifiable genetic cause found otherwise.The paper talks about the treatment of this patient with testosterone replacement and then reviews the limited animal data on methylphenidate and its effects on spermatogenesis. Of course, as with any case report and retrospective review, it is difficult to say for certain whether this is associated or there is another unknown confounding factor involved, so the authors suggestion that reproductive harm has to be taken into account before giving methylphenidate to children is premature at best.However, this case does highlight the real dearth of information and studies we have on medications that are widely used and their effects on sperm production and reproductive potential. While this drug is so widely prescribed we only have a number of rat studies and a primate study, all of which have limitations in their ability to be generalized to human results.As a result, this is a great case report for generating ideas, and raising the idea that maybe some things we commonly prescribe can cause reproductive effects we are unsure of, but to say there is an association, and then say that methylphenidate should be prescribed less because of this is a bit of a stretch.", "responses": [ { "c_id": "958", "date": "03 Sep 2014", "name": "Ranjith Ramasamy", "role": "Author Response", "response": "We agree with the reviewer that the association between chronic ritalin use and testicular failure is not direct. The case report is at best hypothesis generating but given some of the strong evidences available in animal studies, the practitioner should be cognizant of a \"possible\" association between chronic ritalin use and infertility and delayed puberty." } ] }, { "id": "6064", "date": "08 Sep 2014", "name": "Robert Matthew Coward", "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 case report of an unrecognized, but only possible, association between use of Ritalin® and delayed puberty. The discussion offers a very nice commentary on the available literature, and this paper overall generates appropriate consideration for this potential serious adverse effect of methylphenidate. The patient reported had almost lifelong use of methylphenidate throughout childhood and adolescence, but stopped several years prior to presenting with delayed puberty. Another reviewer made several very important comments that I was going to make, so I will not repeat them. A few additional considerations to improve this very nice manuscript are discussed below:This patient's particular situation was unusual in that he presented with normal gonadotropins along with hypogonadism, and this should be discussed. Delayed puberty is not typically a testicular problem, but rather one with the HPG axis. The hypothesis would be that the methylphenidate has a direct suppressive effect at the level of the testicles. The combined use of exogenous testosterone and hCG has been reported by Hsieh et al. to maintain spermatogenesis in hypogonadal men, but the combination has not been previously reported to initiate puberty. It certainly would be the most physiologic manner to initiate puberty, and this point could be highlighted. The options for treatment should be discussed. These can include several options such as exogenous testosterone, high dose hCG, or a combination of the two using a lower dose of hCG which was ultimately chosen. Please include the dose of Testopel used. The androgenization of the patient is noted through the use a exogenous testosterone. An improvement in testicular function is not documented by the authors. This can include post-therapy changes in testicular volume, or more accurately a post-therapy semen analysis.\n\nIn the Follow-up and Outcomes section, the sentence beginning, \"As for his fertility...\" falls flat with mere speculation. A brief discussion of the potential benefit from ultimately stopping exogenous testosterone and reversing its effects with higher dose hCG is warranted, as it is unclear from this report that the patient has irreversible testicular atrophy and hypofunction that would result in long-term non-obstructive azoospermia requiring testicular sperm extraction. A simple follow-up semen analysis would be the most accurate way to update the manuscript and frame this discussion.", "responses": [] }, { "id": "6011", "date": "11 Sep 2014", "name": "Jay I. Sandlow", "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 OF STUDY The authors present a case report in which there is an association between chronic methylphenidate use from a very young age and delayed puberty and reproductive growth. The use of the term “testicular failure” is a misnomer. His low gonadotropins make this secondary testicular failure. There is no evidence that he has primary testicular failure, as the authors do not present any data to show lack of response to gonadotropins. At best, this may be an association. SPECIFIC COMMENTS FOR AUTHORSIntroductionThe introduction should conclude with a restatement of the goal or thesis of the manuscript. It currently ends describing mice fertility.CasePlease clarify for the semen analysis “after use of pellet”. Do the authors mean after the examination of the spun down SA pellet? If so please rephrase this.Was delayed puberty ever ruled out in this patient? Was he examined by a pediatric endocrinologist?Why was the patient not started on HCG alone to see what his testicular response would be? By giving him exogenous testosterone, his gonadotropins will remain low and his spermatogenesis suppressed.In follow up, please describe more clearly the upgrading in his Tanner stageDiscussionThe second sentence of the discussion is much too strong. To this point, the authors have provided no compelling evidence of the relationship between chronic Ritalin use and reproductive function. Only midway through the discussion do the authors describe this more as an association than a causal mechanism. Clarify at the outset that there is no human data suggesting this relationship currently available.For the rat studies, did these rats have exposure prior to adolescence and signs of maturation? Only in this way can the rats possibly be compared to this human and serve as a more direct causal possibility.Reference 10, as the only human study cited, should be explored more than the rat and monkey studies. It should probably also be described earlier in this case report.The authors should state whether or not there is data from the methylphenidate approval trials that demonstrates suggestion of delayed puberty or delayed development among prepubertal boys taking this medication.In the concluding paragraph, the authors again should consider tempering their comments, that this is at most an association and not a direct causal relationship.", "responses": [] } ]	1	https://f1000research.com/articles/3-207
https://f1000research.com/articles/3-168/v1	22 Jul 14	{ "type": "Research Article", "title": "The effects of extremes of pH on the growth and transcriptomic profiles of three haloarchaea", "authors": [ "Aida Moran-Reyna", "James A. Coker", "Aida Moran-Reyna" ], "abstract": "The halophilic archaea (haloarchaea) live in saline environments which are found across the globe. In addition to salinity, these niches can be quite dynamic and experience extreme conditions such as low oxygen content, radiation (gamma and UV), pH and temperature. However, of all the naturally occurring stresses faced by the haloarchaea, only one, pH, has not been previously reported on. Therefore, we endeavored to determine the responses of the transcriptomes of three haloarchaea (Hla, Hvo, and NRC-1) to growth under acidic and alkaline pH. Our observations showed that the transcriptomes of Hvo and NRC-1 respond in a similar manner to each other as well as other prokaryotes when grown in an acidic environment, while the pattern for Hla was dissimilar. For alkaline stress, all three haloarchaea responded in a manner similar to well-studied archaea and bacteria and had four-times more significantly regulated transcripts in common, compared to acidic growth. Additionally, we performed an analysis on the changes in the transcriptomes of the three haloarchaea when shifting from one pH extreme to the other. The results showed that the transcriptomes of all three haloarchaea respond more similarly when moving from alkaline to acidic conditions compared to moving from an acidic to alkaline environment. Interestingly, our studies also showed that individual genes of multiple paralogous gene families (tbp, tfb, orc/cdc6, etc.) found in the haloarchaea were regulated under specific stresses thereby providing evidence that they modulate the response to various environmental stresses. The studies described here are the first to catalog the changes in the haloarchaeal transcriptomes under growth in extreme pH and help us understand how life is able to thrive under all conditions present on Earth and, if present, on extraterrestrial bodies as well.", "keywords": [ "The halophilic archaea (haloarchaea) live in saline environments", "such as the Great Salt Lake in Utah", "the Dead Sea", "and solar salterns. They belong to a single order (Halobacteriales)", "which consists of a single family (Halobacteriaceae)", "making them genetically similar to each other1. Many haloarchaea are facultative aerobes as extreme saline environments often have low oxygen solubility2. Haloarchaea have adapted to survive in their saline environments by employing a tactic called “salting in” whereby they selectively uptake K+ and Cl- ions to a concentration of greater than three molar. To keep their proteins from precipitating in these high ionic strength conditions", "haloarchaea have also adapted a primarily acidic proteome with proteins containing most of their negative charges on their surface3. The negative charges on the surface help orient water molecules to keep the proteins hydrated thereby stopping the possibility of precipitation." ], "content": "Introduction\n\nThe halophilic archaea (haloarchaea) live in saline environments, such as the Great Salt Lake in Utah, the Dead Sea, and solar salterns. They belong to a single order (Halobacteriales), which consists of a single family (Halobacteriaceae), making them genetically similar to each other1. Many haloarchaea are facultative aerobes as extreme saline environments often have low oxygen solubility2. Haloarchaea have adapted to survive in their saline environments by employing a tactic called “salting in” whereby they selectively uptake K+ and Cl- ions to a concentration of greater than three molar. To keep their proteins from precipitating in these high ionic strength conditions, haloarchaea have also adapted a primarily acidic proteome with proteins containing most of their negative charges on their surface3. The negative charges on the surface help orient water molecules to keep the proteins hydrated thereby stopping the possibility of precipitation.\n\nAlthough saline environments are often studied in relation to salt content, they have been shown to have other extreme conditions: temperature, gamma radiation, oxygen sensitivity, UV, and pH, which all affect the adaptation and survival of the species in these niches. Of these conditions, temperature4, salinity4, oxygen requirements5, and radiation6,7 are well studied. For extremes of temperature, it has been previously shown that haloarchaea over-express chaperones (hsp and csp), DNA binding proteins (hpyA), and other previously characterized temperature stress genes/proteins8. For high energy (gamma) radiation, the haloarchaea have been shown to be the second most naturally resistant life form on the planet, slightly less resistant than Deinococcus radiodurans7. They are also able to withstand UV intensities about 100× higher than those experienced by organisms on the surface of Earth (>250 J/m2 vs. 2–3 J/m2)6,9,10. Haloarchaea are able to survive these conditions through an overexpression of DNA binding/repair proteins7,9–11.\n\nAlthough previous studies of the haloarchaea have not focused on pH, the response from other prokaryotic organisms such as Escherichia coli and Bacillus subtilis has been thoroughly examined12–14. These studies are important because of the role that pH plays in disease and its influence on the passage of prokaryotic species through and/or colonization of specific areas of the human body (e.g. acidic conditions in the stomach and beginning of the small intestine and alkaline conditions in the intestines). E. coli, when the environment is acidic, tends to use pathways that consume acids as well as increase the production of the inner membrane protein YagU and hydrogenases; however, when the environment is alkaline metabolic genes are up-regulated as are pathways that result in the production of acids to neutralize the effect of alkaline pH in the external environment12,14. For B. subtilis, acidic stress is of extreme importance because it leads to up-regulation of spore germinating genes, which allow the organism to survive in the hostile conditions of the stomach13,15.\n\nThe best studied extremophilic organisms in regard to the effects of pH are the acidophiles and the alkaliphiles. These studies have primarily shown that these organisms survive via passive and active responses. Passive responses involve the modification of the cell wall/membrane to keep out excess H+ and OH- ions16,17. Active responses primarily involve pumping in/out specific ions. For alkaliphiles this is accomplished with the use of Na+/H+ antiporters where Na+ is pumped out of the cell and H+ is pumped in, as well as an up-regulation of pathways that produce acid16; however, the active processes used by acidophiles are not well known.\n\nInsights into the mechanisms of archaeal adaptations to the extremes of high salinity2 and acidity18 or alkalinity19 have been gained via genome sequencing projects by addressing each issue separately. However, more in depth knowledge of dual extremes is lacking. Haloarchaea are one of the few groups of Archaea that have been isolated from both extremes of the pH scale19,20 and therefore represent a novel group of organisms that are ideal subjects for studies to determine the transcriptomic responses from multiple extremes. On the lower end of the pH scale, haloarchaea are commonly found in acidic lakes such as Lake Afrera, Ethiopia, and Lake Aerodrome and Lake Brown in Western Australia20. On the higher end of the pH scale, the haloarchaea are commonly found in alkaline lakes such as Lake Natron and Lake Magadi in the Great Rift Valley in Africa as well as Mono Lake in California19,21.\n\nTherefore, we undertook the following study to determine the changes in transcriptomes of three well-studied haloarchaea: Halorubrum lacusprofundi, Haloferax volcanii, and Halobacterium sp. NRC-1, to understand the effects on the transcriptome associated with growth of haloarchaea at extremes of pH. These organisms were selected for their ability to grow in extreme saline conditions, as well as having a fully sequenced genome, and because they can be easily grown and manipulated within the laboratory. Broadly, we found that under acidic conditions the response of the transcriptomes of Hvo and NRC-1 were similar and also paralleled the responses seen in E. coli, while the response from Hla appears to be species-specific. For alkaline conditions, the response of the transcriptomes from all three haloarchaea contained four-times the number of regulated transcripts compared to the acidic conditions. The shared response also paralleled that seen in E. coli and B. subtilis as well as alkaliphiles. Finally, we performed an analysis of shifts from one pH extreme to the other. This analysis showed that the shift from alkaline to acidic conditions produced a more similar response in all three organisms compared to the reverse. The response seen in our analyses also showed that several important genes were regulated under stress including orc/cdc6 and tfb. This result was quite interesting as it is an example of one gene with multiple paralogs showing a response under environmental stress, thereby providing evidence for the hypothesis22,23 that these multiparalogous genes in the haloarchaea are used to respond to and modulate the response to various environmental stresses.\n\n\nMaterials and methods\n\nHalorubrum lacusprofundi (Hla) cultures were grown in standard Artificial Deep Lake medium at 30°C24. Haloferax volcanii (Hvo) cultures were grown in standard HV-YPC medium at 49°C25. Halobacterium sp. NRC-1 (NRC-1) cultures were grown in standard CM+ medium at 42°C2. Each organism was grown in an Innova 42R platform shaker at 220 rpm. For growth in varying pH, the above media were prepared as described but the pH was altered to 4.4, 5.4, 6.4 and 8.4 for Hla, 4.5, 5.5, 6.0, 6.5, 8.0 and 8.5 for Hvo, and 4.2, 5.2, 6.2, 8.2 and 9.2 for NRC-1. Growth curves were measured in 50 mL cultures by removing 1 mL aliquots at various time points and measuring the optical density (OD600) in a Shimadzu UV-160 spectrophotometer. The doubling times of each organism, under each pH condition, were derived by calculating the line of least squares during the exponential phase of growth. Each growth curve was measured twice and replicated once.\n\nOligomer (60-mer) probes used in our arrays were designed using the program PICKY (http://www.complex.iastate.edu) with the following criteria: selection_left_begin = 0, selection_right_end = 0, maximum_oligo_size = 70, minimum_oligo_size = 60, maximum_match_len = 15, minimum_match_len = 13, maximum_gc_content = 70, minimum_gc_content = 50, candidates_per_gene = 5, probes_per_gene = 4, minimum_similarity = 75, minimum_temp_separation = 10. For each organism, the arrays contained 13 probes for each of the 3538 annotated genes in Hla, eight probes for each of the 4020 genes in Hvo, and 17 probes for each of the 2524 genes in NRC-1. Oligonucleotide arrays were in situ synthesized by Agilent using ink-jet technology and used for transcriptome analysis of all three organisms. Signal intensities with a dynamic range in excess of three orders of magnitude were found allowing simultaneous analysis of low and high-intensity features.\n\nFor each microarray experiment, two cultures were grown in the same manner as those used for the growth curves. Cultures for all three organisms were harvested at late exponential phase (OD600 = 0.9–1.0) for microarray analysis. Total nucleic acids (DNA and RNA) were purified from the cultures using the Agilent total RNA isolation mini kit. Total nucleic acids were then incubated with RNase-free DNase I (New England Biolabs) to digest the genomic DNA. RNA was then purified using the Agilent total RNA isolation mini kit. RNA was pooled from both cultures for cDNA synthesis. cDNA was prepared for control and experimental samples using fluorescently labeled Cy3-dCTP and Cy5-dCTP, respectively. Labeled cDNAs were sent to the Interdisciplinary Center for Biotechnology Research (ICBR) at the University of Florida for hybridization and scanning. Concentrations of cDNA were measured using a Nanodrop (ND-1000) spectrophotometer (NanoDrop Technologies, Wilmington, DE) at ICBR. Hybridization was performed as recommended by Agilent and as previously described8,26 using 1 µg (0.5 µg for experimental and control) of total cDNA per microarray. Microarray analysis was performed in duplicate for technical replicates and slides were scanned for Cy-3 and Cy-5 signals with an Agilent DNA-microarray scanner as previously done5,23.\n\nProbe signals were extracted and initial analysis was done with the Agilent Feature Extraction Software, where signal from each channel was normalized using the LOWESS algorithm to remove intensity-dependent effects within the calculated values. The data were further parsed using Agilent’s Genespring GX software and Perl scripts, written by Dr. James A. Coker, which calculated various statistical measurements. Genes showing greater than 2-fold change in transcript abundance in at least two of the replicates with an illuminant intensity greater than 7 and less than 14 were selected for further analysis, as described previously8,26,27.\n\n\nResults and discussion\n\nWe employed three well-studied haloarchaea: Halorubrum lacusprofundi, Haloferax volcanii, and Halobacterium sp. NRC-1, to understand the effects that pH stress plays on the transcriptome of haloarchaea. These organisms were selected due to the depth of knowledge in relation to adaption/survival to other naturally occurring stresses8,28–31. This element was key to tease out differences between the general stress response and pH specific responses. We also chose three organisms, as opposed to one, to ascertain if the response to pH was specific to the organism, the Halobacteriaceae, the Prokaryotes, or the three Domains of Life.\n\nThe growth optimum for the Halobacteriaceae is around pH 72,24,25. Specifically, for the organisms used in this study, it is pH 7.4, 7.5, and 7.2 for Halorubrum lacusprofundi, Haloferax volcanii, and Halobacterium sp. NRC-1, respectively. In order to determine the changes in gene transcripts during growth at pH extremes we first determined the range of growth for each organism as it was previously unknown.\n\nHalorubrum lacusprofundi – Growth was observed from pH 6.4 to 8.4 (Figure 1). The doubling time at pH 6.4 was around 64 hours and at 8.4 was 77 hours, both well above the 44 hours at pH 7.4 (optimum). Growth was tested but not observed at pH 4.4 or 5.4.\n\nGrowth of Halorubrum lacusprofundi at pH 4.4 (black ×s), pH 5.4 (light green circles), pH 6.4 (light blue diamonds), pH 7.4 (red squares), and pH 8.4 (blue diamonds).\n\nHaloferax volcanii – Growth was observed from pH 6.0 to 8.0 (Figure 2). Doubling time at pH 6.0 was about 7.0 hours while at pH 6.5 it almost halved to about 3.7 hours. At the typical laboratory growth pH of 7.5 the doubling time was about 3.2 hours. Under alkaline growth conditions (pH 8.0), the doubling time was about 4.0 hours. Growth was tested at pH 4.5, 5.5 and 8.5 but not observed.\n\nGrowth of Haloferax volcanii at pH 4.5 (black asterisks), pH 5.5 (black ×s), pH 6.0 (light green circles), pH 6.5 (light blue diamonds), pH 7.5 (red squares), pH 8.0 (blue diamonds), and pH 8.5 (green circles).\n\nHalobacterium sp. NRC-1 – Growth was observed from pH 5.2 to 9.2 (Figure 3). Doubling times at pH 8.2, 7.2 and 6.2 were about 8.6, 8.6, and 9.6 hours, respectively. Growth at pH 5.2 was considerably slower and had a doubling time around 75 hours. Growth at pH 9.2 was the slowest recorded with a doubling time of about 150 hours. Growth was tested but not observed at pH 4.2.\n\nGrowth of Halobacterium sp. NRC-1 at pH 4.2 (black Xs), pH 5.2 (light green circles), pH 6.2 (light blue diamonds), pH 7.2 (red squares), pH 8.2 (blue diamonds), and pH 9.2 (green circles).\n\nOnce the pH range of growth was determined, we grew each of the three organisms at its optimum pH and at the most acidic and alkaline pH where growth was observed. This was done in an effort to maximize the potential differences (if any) in the transcriptomes. We designed all three microarrays (one per organism) used for these experiments and each incorporated at least eight probes per gene in the entire genome. Hybridizations were designed so a positive fold-change value would correspond with an up-regulation of a gene transcript in the experimental condition (growth in acidic or alkaline conditions) and a negative value would correspond with a down-regulation of a gene transcript.\n\nHalorubrum lacusprofundi – For H. lacusprofundi, microarray analysis showed that transcripts from 532 genes were significantly up-regulated and 608 were down-regulated at pH 8.4, compared to pH 7.4 (Table 1). For the up-regulated gene transcripts, the largest percent were members of COG group E (amino acid transport and metabolism group) and the lowest were COG group D (cell cycle control, cell division, chromosome partitioning). For the down-regulated gene transcripts, the largest percent were members of COG group R (general function prediction only) and the lowest was COG group M (cell wall/membrane/envelope biogenesis). About 65% of the up-regulated and 45% of the down-regulated transcripts in Hla were from genes without an assigned function (no COG, general function, and unknown function), which is typical for the Archaea.\n\nDespite the large number of unknowns, there were several transcripts/genes of interest in Hla. Of these, most of the up-regulated transcripts fell into the following groups: tRNA, metabolism, and stress genes. For the tRNA transcripts, all 27 showed an up-regulation from 2- to 11-fold. For metabolism transcripts, these included phosphoenolpyruvate (PEP) carboxylase and cytochrome P450, with its corresponding ferridoxin. For the stress genes, transcripts of genes such as the small heat shock (hsp20 family), Cpn60/TCP-1, and universal stress (uspA) were regulated. For the down-regulated transcripts, they mainly belonged in the following categories: translation, stress response, and dehydrogenases. For the translation transcripts, 40% were from ribosomal proteins. There were also two tRNA sythetases (alanine and lysine), two tRNAs (glycine and valine) and a tRNA methyltransferase that were down-regulated. Fifteen dehydrogenase transcripts, which are involved in various aspects of metabolism from lipid metabolism, energy conversion, and amino acid transport, were also down-regulated. Interestingly, the transcript from an aminoglycosidase phosphotransferase (Hlac_0222), which is commonly annotated as a main mechanism of kanamycin resistance32, was also down-regulated. Hlac_0222 was one of the most down-regulated transcripts (-58 fold) in Hla grown in alkaline conditions. It is puzzling why an antibiotic resistance gene would be down-regulated under pH stress; however, it seems clear that the lack of this gene product is important during growth at alkaline pH. While little is know about Hlac_0222, it does have two conserved domains that belong to kinase (cd05154) and phosphotransferases (pfam01636). Therefore, it is possible that Hlac_0222 acts as a general regulator through the transfer of phosphate groups under growth at non-alkaline conditions while this function is down-regulated during growth under alkaline conditions. Transcriptional regulators of the asnC, iclR, padR, tetR, and XRE families were also down-regulated.\n\nHaloferax volcanii – For H. volcanii, microarray analysis showed that transcripts from 326 genes were significantly up-regulated and 1581 were down-regulated at pH 8.0 compared to pH 7.5 (Table 1). For the up-regulated gene transcripts, the largest percent were members of COG group E (amino acid transport and metabolism) and the lowest percent were members of COG group U (intracellular trafficking, secretion, and vascular transport) and N (cell motility). For the down-regulated gene transcripts, the largest percent were members of COG group E (amino acid transport and metabolism) and the lowest percent were members of COG group A (RNA processing) and B (chromatin structure and dynamics).\n\nAs with Hla, the predominant group of up-regulated (58%) and down-regulated (47%) transcripts was from genes without an assigned function (no COG, general function, and unknown function). For the up-regulated transcripts, those with an annotated function mainly grouped as follows: tRNA, metabolism, and stress. For the tRNAs, 31 were regulated from 2- to 1900-fold. For the stress related genes, the small heat shock (hsp20 family) and universal stress (uspA) were regulated as was observed in Hla. Additionally, transcripts of the groEL chaperone were also up-regulated. For the metabolism, transcripts such as triosephosphate isomerase and beta-glucosidase were significantly up-regulated. For the down-regulated with an assigned function, transcripts from 54 genes are from dehydrogenases, which act in a variety of roles within the cells. Transcripts related to translation were also down-regulated including several tRNA synthetases, tRNA genes, and aminotransferases. Several general transcription factors were also down-regulated including one TATA-binding protein (TBP) and six transcription factor B proteins (TFBs).\n\nHalobacterium sp. NRC-1 – For Halobacterium sp. NRC-1, microarray analysis showed that transcripts from 471 genes were significantly up-regulated and 72 were down-regulated at pH 8.2 compared to pH 7.2 (Table 1). For the up-regulated gene transcripts, the largest percent were from COG group J (translation, ribosome structure and biogenesis) and the lowest percent were from COG group U (intracellular trafficking, secretion, and vascular transport). For the down-regulated gene transcripts, the largest percent were from COG group P (inorganic ion transport and metabolism) and H (coenzyme transport and metabolism) and the lowest percent were from COG group C (energy production and conversion) and I (lipid transport and metabolism).\n\nFor NRC-1, like Hla and Hvo, the majority of the up-regulated (38%) and down-regulated (61%) transcripts did not have an annotated function. However, for those with an annotation, most of the up-regulated genes were involved in ribosome formation, metabolism and stress. Metabolism genes up-regulated in NRC-1 included aconitase, several cytochromes (b6 and c oxidase), and citrate synthase. Stress genes included superoxide dismutase, recA, rpa, uspA, and groEL. For the down-regulated transcripts, they primarily included dehydrogenases. However, the transcript from vng0070h is also of note as it was the most down-regulated transcript (-300 fold). The corresponding gene is annotated to be involved in plasmid stability (COG3668). However, this gene is also a member of COG2026, which contains RelE, which is known to cleave mRNA in the A-site in ribosomes.\n\nIn summary, under growth in alkaline conditions, the three haloarchaea studied (Hla, Hvo, and NRC-1) had transcripts from 36 and 15 unique genes (not counting paralogs) in common for the up and down-regulated, respectively (Dataset 3). This is based on the fact that the genes/transcripts were members of the same COG and/or pfam. Taken together these results show the coordination of 51 unique genes in the same direction (up- or down-regulated) across all three organisms. Specifically, 1.4% of Hla, 1.3% of Hvo, and 2.0% of NRC-1 genes are regulated in the same fashion during growth in alkaline pH.\n\nA majority of these shared transcripts code for products involved in either metabolism (glucose 1-dehydrogenase, malate dehydrogenase, succinate dehydrogenase) or chaperones (small heat shock protein, universal stress protein, Cpn60/TCP-1). Another transcript of note shared by all three organisms belongs to COG 1405 (transcription initiation factor IIB). It has been hypothesized26,33 that TFBs of archaea act in a similar fashion to the sigma factors of bacteria. Therefore this TFB, most similar to TfbBF in NRC-126, regulated in all three organisms, may be the general transcription factor responsible for regulating the response to alkaline pH stress. As seen with B. subtilis14 and E. coli13 and alkaliphiles16, all three organisms predominantly up-regulated genes involved with metabolism and those that produce acid in response to an alkaline environment.\n\nHalorubrum lacusprofundi – For H. lacusprofundi, microarray analysis showed that 25 gene transcripts were significantly up-regulated and 43 were down-regulated at pH 6.4, compared to pH 7.4 (Table 2). For the up-regulated gene transcripts, the largest percent were members of COG group P (inorganic ion and metabolism) and the lowest were members of GOG groups E (amino acid transport and metabolism) and O (post-translational modification, protein turnover and chaperones). For the down-regulated gene transcripts, the largest percent were members of GOG group E (amino acid transport and metabolism) and the lowest were members of GOG group C (energy production and conversion), K (transcription), and T (signal transduction mechanism).\n\nThe majority of the up-regulated (36%) and down-regulated (34%) transcripts did not have an annotated function. However, there were still two genes of interest: hlac_3059 and hlac_3556, which are homologs of exsB and spoVT. The genes that produce these transcripts are homologs of genes in B. subtilis that are up-regulated during periods of spore formation and dormancy34,35. Hla has not been observed to form spores24, so the regulation of these two genes suggests that the organism is becoming dormant in response to growth in acidic conditions. As with Hvo and NRC-1 (described below), the down-regulated transcripts from Hla were primarily from unknown genes but those with an annotation were primarily from ABC transporters.\n\nHaloferax volcanii – For H. volcanii, microarray analysis showed that 869 gene transcripts were significantly up-regulated and 987 were down-regulated at pH 6.0 compared to pH 7.5 (Table 2). For the up-regulated gene transcripts, the largest percent were from COG group E (amino acid transport and metabolism) and the lowest percent were from COG group A (RNA processing and modification) and COG group B (chromatin structure and dynamics). For the down-regulated gene transcripts, the largest percent of transcripts were from COG group E (amino acid transport and metabolism) and the lowest from COG group B (chromatin structure and dynamics).\n\nThe largest group of the up-regulated (33%) and down-regulated (46%) transcripts was those without an annotated function. However, for those with an annotation, most of the up-regulated transcripts/genes coded for products involved with stress and motility (e.g. cct2, cct3, cspD3, cheDFR, flaD2J, and arcR1). In Hvo, the cct genes form the thermosome, which is a chaperone complex found in archaeal and eukaryotic cells36. It is known for its ability to refold proteins in the cytosol of these organisms. It is also of note that cct1 was not up-regulated in response to growth in acidic conditions, suggesting that the cct2 and cct3 genes are preferentially used to respond to stress associated with growth in acidic conditions. The up-regulation of the cspD3 transcript is also of note for a similar reason as Hvo encodes four paralogous genes. The genes involved with motility: chemotaxis and flagella, were also up-regulated, suggesting that Hvo has a chemotactic response to acidic pH. The up-regulation of the arcR1 transcript is interesting as it was up-regulated 7-fold. This is a significant change for a regulator, especially one involved in the control of the arginine metabolism (fermentation) pathway37. Additionally, members of the electron transport chain were also up-regulated, suggesting that a switch in metabolism and energy production is being observed in Hvo during growth in acidic conditions. For the down-regulated transcripts of genes with an annotation, most were transporters, primarily of the ABC-type. ABC-type transporters have been shown to be important in a general stress response in Listeria monocytogenes38 as well as in regulation of pH within eukaryal cells and their organelles39. Therefore the down-regulation of these transcripts in Hvo shows that this response is common across all three Domains of Life.\n\nHalobacterium sp. NRC-1 – For Halobacterium sp. NRC-1, microarray analysis showed that 664 gene transcripts were significantly up-regulated and 207 were down-regulated at pH 5.2 compared to pH 7.2 (Table 2). For the up-regulated gene transcripts, the largest percent were members of COG group J (translation, ribosomal structure and biogenesis). The lowest percent of transcripts were members of COG group B (chromatin structure and dynamics). For the down-regulated gene transcripts, the largest percent of transcripts were from COG group L (replication, recombination, and repair) and the lowest percent were from COG group M (cell wall, membrane and envelope biogenesis).\n\nThe largest group of the up-regulated (24%) and down-regulated (57%) transcripts was that without an annotated function. Many of the up-regulated transcripts in NRC-1, from a gene with an annotated function, coded for products involved with stress, hydrogenases, and motility. Some examples of the up-regulated transcripts were cctA, cctB, cspD2, dnaK, ndhG1-5, cheBC1RW2Y, and flaA1A2B1B2B3. The cctA and cctB genes comprise the thermosome and may be acting in a similar manner as in Hvo to ensure proper protein folding under acidic growth conditions. The cspD2 gene/transcript is of great interest because it was up-regulated five times more than cspD1. Therefore, the difference in cspD regulation might suggest that cspD2 is specific for acidic stress. The dnaK gene is also of interest as its product stabilizes misfolding proteins before being shuttled to the GroEL complex40. The up-regulation of the ndhG1-5 genes is also interesting, as an increase in hydrogenases has previously been linked to the acidic pH stress response in E. coli and B. subtilis12–14. The up-regulation of genes involved with motility: chemotaxis (cheB, C1, R, W2, Y) and flagella (flaA1, A2, B1, B2, B3) is also of note as it suggests NRC-1 has a chemotactic response to acidic pH. For the down-regulated transcripts with an annotation, many were transporters such as trkA1, A2 and ABC transporters (vng795 - a yufN homolog). The trk genes are responsible for K+ uptake2 and it is possible that the increase in ionic strength associated with the increase in H+ concentration mimics the high salt of the cell enough for it to slow the uptake of K+41.\n\nIn summary, the responses of Hla, Hvo, and NRC-1 under growth in acidic conditions are quite different to those under alkaline conditions (Datasets 2, 3, 6–7, 12–15, 18–21). Primarily, there does not seem to be as large of a haloarchaeal specific response. There were six up-regulated and seven down-regulated transcripts in common for all three organisms during growth in acidic conditions (Dataset 2). For the up-regulated transcripts, they fell into COGs 0614 (ABC-type F3+ hydroxamate transport system) and 0720 (6-pyruvoyl-tetrahydropterin synthase) and pfam 0005 (ABC-transporter), 00528 (transporter system), 01497 (periplasmic binding protein), and 01609 (transposase). For the down-regulated transcripts, they fell into COGs 0589 (universal stress protein A), 1028 (short-chain family oxidoreductase), 1522 (Lrp transcriptional regulator), pfam 00005 (ABC-transporter), 00106 (short-chain dehydrogenase), 00561 (3-oxoadipate enol-lactone hydrolase/4—carboxymuconolactone decarboxylase), and 00582 (universal stress protein family).\n\nA closer look at the response of the three organisms showed that the changes in the transcriptomes of Hvo and NRC-1 were far more similar to each other than either was to Hla (Dataset 6, 8, 12, 14, 18, 20). Hvo and NRC-1 shared 213 up-regulated transcripts and 55 down-regulated transcripts. For up-regulated transcripts, stress (cct and cspD) and motility (che and fla) genes were in common. This response closely parallels what has been reported for E. coli where heat shock genes were up-regulated in acidic conditions and chemotaxis/flagellar genes down-regulated under alkaline conditions42. For down-regulated transcripts, ABC-transporters were the most common. This shared response accounts for 268 genes or 6.6% of the Hvo and 10.6% of the NRC-1 genes in the respective genomes being regulated in the same fashion.\n\nIn addition to comparing the extremes of pH to the optimal, we also looked at the changes in transcript abundance from the extremes of pH for each organism. Briefly, this comparison generates a ratio such that a positive value indicates an increase in transcript abundance under a shift from alkaline to acidic conditions and a negative number indicates an increase in transcript abundance under a shift from acidic to alkaline conditions. This comparison is fairly common and has been done routinely for E. coli and B. subtilis13,15,42.\n\nAcidic to alkaline – This comparison showed that 20 transcripts (Table 3) that undergo significant changes in a shift from acidic to alkaline conditions are shared across all three haloarchaea studied (Dataset 1). Some of the more interesting transcripts that were regulated in common were one general transcription factor (related to tfbE), a cell wall/membrane, ABC-transporters, and one replication origin recognition (orc/cdc6). The general transcription factor is of interest as it has been hypothesized that the Tfbs play a primary role is regulating genes, similar to the sigma factors of bacteria26. Therefore, the data strongly suggests that the tfbE gene is important in the response in all three organisms when the environmental conditions move from acidic or neutral to alkaline conditions. This is the first evidence from an environmental study in support of this more than ten year-old hypothesis. The orc/cdc6 gene is of interest for similar reasons. In NRC-1 this gene (orc1) is non-essential; however, like the general transcription factors, the multiple orc paralogs found in every haloarchaeon are hypothesized to play a role in alternate start site recognition23. Another interesting note is that a transcript from another orc gene (orc5) was regulated under acidic growth possibly providing evidence for alternative replication start sites when growth occurs under pH stress. The ABC-transporters and wall/membrane genes are important, as they possibly constitute a conserved mechanism of the haloarchaea for surviving alkaline conditions, or the shift to alkaline conditions, by a mechanism involving active transport.\n\nAlkaline to acid – For a shift from alkaline to acidic conditions we observed that 135 of the transcripts (Table 3 and Dataset 1) are shared across all three organisms. Some of the more interesting genes that were regulated in common were an orc/cdc6 gene, several ribosomal genes, and a spermidine ABC-transporter. The orc/cdc6 gene is of interest as this gene in NRC-1 (orc5) is non-essential; however, like the general transcription factors, the multiple orc paralogs found in every haloarchaeon are hypothesized to play a role in alternate start site recognition. The orc5 gene is located on the main chromosome in all three organisms and might play a role in alternative replication start sites under stress conditions. The spermidine ABC-transporter is of interest as spermidine is used as a pH regulator in cells to maintain homeostasis. Its up-regulation in acidic environments suggests that the haloarchaea are responding in a manner similar to many other cells43 from the other domains of life and possibly constitute a conserved mechanism for acidic survival in all three Domains of life.\n\n\nData availability\n\nF1000Research: Dataset 1. Data of pH extremes effects on haloarchaea growth and transcriptomic profiles, 10.5256/f1000research.4789.d3254444.", "appendix": "Author contributions\n\n\n\nAMR and JAC conceived the study. JAC designed the experiments. AMR carried out the research. JAC provided expertise on microarray design. AMR prepared the first draft of the manuscript. Both authors were involved in the revisions of the drafts 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 start up funds given to James A. Coker from the University of Alabama at Birmingham.\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 all the members of the Coker lab, especially Ryne Black, for their help and the Watts lab, especially Mickie Powell, for allowing us to use their spectrophotometer. Finally, we would like to thank our families for all their support.\n\n\nReferences\n\nDasSarma S, Coker JA, DasSarma P: in Encyclopedia of Microbiology, M. Schaechter, Ed. (Elsevier, 2009), vol. 3rd, 1–23.\n\nNg WV, Kennedy SP, Mahairas GG, et al.: Genome sequence of Halobacterium species NRC-1. Proc Natl Acad Sci U S A. 2000; 97(22): 12176–12181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKennedy SP, Ng WV, Salzberg SL, et al.: Understanding the adaptation of Halobacterium species NRC-1 to its extreme environment through computational analysis of its genome sequence. Genome Res. 2001; 11(10): 1641–1650. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoker JA, DasSarma P, Kumar J, et al.: Transcriptional profiling of the model Archaeon Halobacterium sp. NRC-1: responses to changes in salinity and temperature. Saline Systems. 2007; 3: 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMüller JA, DasSarma S: Genomic analysis of anaerobic respiration in the archaeon Halobacterium sp. strain NRC-1: dimethyl sulfoxide and trimethylamine N-oxide as terminal electron acceptors. J Bacteriol. 2005; 187(5): 1659–1667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaliga NS, Bjork SJ, Bonneau R, et al.: Systems level insights into the stress response to UV radiation in the halophilic archaeon Halobacterium NRC-1. Genome Res. 2004; 14(6): 1025–1035. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeVeaux LC, Muller JA, Smith J, et al.: Extremely radiation-resistant mutants of a halophilic archaeon with increased single-stranded DNA-binding protein (RPA) gene expression. Radiat Res. 2007; 168(4): 507–514. PubMed Abstract \| Publisher Full Text\n\nCoker JA, DasSarma P, Kumar J, et al.: Transcriptional profiling of the model Archaeon Halobacterium sp. NRC-1: responses to changes in salinity and temperature. Saline Systems. 2007; 3: 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrowley DJ, Boubriak I, Berquist BR, et al.: The uvrA, uvrB and uvrC genes are required for repair of ultraviolet light induced DNA photoproducts in Halobacterium sp. NRC-1. Saline Systems. 2006; 2: 11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCready S, Müller JA, Boubriak I, et al.: UV irradiation induces homologous recombination genes in the model archaeon, Halobacterium sp. NRC-1. Saline Systems. 2005; 1: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCready S, Marcello L: Repair of UV damage in Halobacterium salinarum. Biochem Soc Trans. 2003; 31(Pt 3): 694–698. PubMed Abstract\n\nMaurer LM, Yohannes E, BonDurant SS, et al.: pH regulates genes for flagellar motility, catabolism, and oxidative stress in Escherichia coli K-12. J Bacteriol. 2005; 187(1): 304–319. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlonczewski JL, Coker JA, DasSarma S, et al.: The microbial habitability zone: surviving with multiple stressors. Microbe. 2009; 5(3): 110–116. Reference Source\n\nWilks JC, Kitko RD, Cleeton SH, et al.: Acid and base stress and transcriptomic responses in Bacillus subtilis. Appl Environ Microbiol. 2009; 75(4): 981–990. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilks JC, Kitko RD, Cleeton SH, et al.: Acid and base stress and transcriptomic responses in Bacillus subtilis. Appl Environ Microbiol. 2009; 75(4): 981–990. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrulwich TA, Ito M, Gilmour R, et al.: Mechanisms of cytoplasmic pH regulation in alkaliphilic strains of Bacillus. Extremophiles. 1997; 1(4): 163–169. PubMed Abstract \| Publisher Full Text\n\nSharma A, Kawarabayasi Y, Satyanarayana T: Acidophilic bacteria and archaea: acid stable biocatalysts and their potential applications. Extremophiles. 2012; 16(1): 1–19. PubMed Abstract \| Publisher Full Text\n\nChen L, Brügger K, Skovgaard M, et al.: The genome of Sulfolobus acidocaldarius, a model organism of the Crenarchaeota. J Bacteriol. 2005; 187(14): 4992–4999. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFalb M, Pfeiffer F, Palm P, et al.: Living with two extremes: conclusions from the genome sequence of Natronomonas pharaonis. Genome Res. 2005; 15(10): 1336–1343. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMormile MR, Hong BY, Benison KC: Molecular analysis of the microbial communities of Mars analog lakes in Western Australia. Astrobiology. 2009; 9(10): 919–930. PubMed Abstract \| Publisher Full Text\n\nKulp TR, Han S, Saltikov CW, et al.: Effects of imposed salinity gradients on dissimilatory arsenate reduction, sulfate reduction, and other microbial processes in sediments from two California soda lakes. Appl Environ Microbiol. 2007; 73(16): 5130–5137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaliga NS, Goo YA, Ng WV, et al.: Is gene expression in Halobacterium NRC-1 regulated by multiple TBP and TFB transcription factors? Molecular Microbiol. 2000; 36(5): 1184–1185. PubMed Abstract \| Publisher Full Text\n\nCoker JA, DasSarma P, Capes M, et al.: Multiple replication origins of Halobacterium sp. strain NRC-1: properties of the conserved orc7-dependent oriC1. J Bacteriol. 2009; 191(16): 5253–5261. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFranzmann PD, Stackebrant E, Sanderson K, et al.: Halobacterium lacusprofundi sp. nov., a halophilic bacterium isolated from Deep Lake, Antarctica. Sys Appl Microbiol. 1988; 11(1): 20–27. Publisher Full Text\n\nAllers T: Overexpression and purification of halophilic proteins in Haloferax volcanii. Bioeng Bugs. 2010; 1(4): 288–290. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoker JA, DasSarma S: Genetic and transcriptomic analysis of transcription factor genes in the model halophilic Archaeon: coordinate action of TbpD and TfbA. BMC Genet. 2007; 8: 61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuackenbush J: Microarray data normalization and transformation. Nat Genet. 2002; 32(Suppl): 496–501. PubMed Abstract \| Publisher Full Text\n\nDasSarma S, Berquist BR, Coker JA, et al.: Post-genomics of the model haloarchaeon Halobacterium sp. NRC-1. Saline Systems. 2006; 2: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoo YA, Yi EC, Baliga NS, et al.: Proteomic analysis of an extreme halophilic archaeon, Halobacterium sp. NRC-1. Mol Cell Proteomics. 2003; 2(8): 506–524. PubMed Abstract \| Publisher Full Text\n\nHartman AL, Norais C, Badger JH, et al.: The complete genome sequence of Haloferax volcanii DS2, a model archaeon. PLoS One. 2010; 5(3): e9605. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoppa J: Functional genomic and advanced genetic studies reveal novel insights into the metabolism, regulation, and biology of Haloferax volcanii. Archaea. 2011; 2011: 602408. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStogios PJ, Spanogiannopoulos P, Evdokimova E, et al.: Structure-guided optimization of protein kinase inhibitors reverses aminoglycoside antibiotic resistance. Biochem J. 2013; 454(2): 191–200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu Q, Han J, Zhou L, et al.: Dissection of the regulatory mechanism of a heat-shock responsive promoter in Haloarchaea: a new paradigm for general transcription factor directed archaeal gene regulation. Nucleic Acids Res. 2008; 36(9): 3031–3042. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcPherson SA, Li M, Kearney JF, et al.: ExsB, an unusually highly phosphorylated protein required for the stable attachment of the exosporium of Bacillus anthracis. Mol Microbiol. 2010; 76(6): 1527–1538. PubMed Abstract \| Publisher Full Text\n\nRamirez-Peralta A, Stewart KA, Thomas SK, et al.: Effects of the SpoVT regulatory protein on the germination and germination protein levels of spores of Bacillus subtilis. J Bacteriol. 2012; 194(13): 3417–3425. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLarge AT, Kovacs E, Lund PA: Properties of the chaperonin complex from the halophilic archaeon Haloferax volcanii. FEBS Lett. 2002; 532(3): 309–312. PubMed Abstract \| Publisher Full Text\n\nReupp A, Soppa J: Fermentative arginine degradation in Halobacterium salinarium (formerly Halobacterium halobium): genes, gene products, and transcripts of the arcRACB gene cluster. J Bacteriol. 1996; 178(16): 4942–4947. PubMed Abstract \| Free Full Text\n\nLiu Y, Ceruso M, Gunther NW IV, et al.: Construction of Listeria monocytogenes Mutants with In-Frame Deletions in Putative ATP-Binding Cassette (ABC) Transporters and Analysis of Their Growth under Stress Conditions. J Microb Biochem Technol. 2012; 4: 7. Reference Source\n\nBrazill DT, Meyer LR, Hatton D, et al.: ABC transporters required for endocytosis and endosomal pH regulation in Dictyostelium. J Cell Sci. 2001; 114(Pt 21): 3923–3932. PubMed Abstract\n\nKim W, Park YR, Im S, et al.: Expression and bioconversion of recombinant m- and p-hydroxybenzoate hydroxylases from a novel moderate halophile, Chromohalobacter sp. Biotechnol Lett. 2012; 34(9): 1687–1692. PubMed Abstract \| Publisher Full Text\n\nElcock AH, McCammon JA: Electrostatic contributions to the stability of halophliic proteins. J Mol Biol. 1998; 280(4): 731–748. PubMed Abstract \| Publisher Full Text\n\nMauer LM, Yohannes E, BonDurant SS, et al.: pH regulates genes for flagellar motility, catabolism, and oxidative stress in Escherichia coli. K-12. J Bacteriol. 2005; 187(1): 304–319. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMangold S, Rao Jonna V, Dopson M: Response of Acidithiobacillus caldus toward suboptimal pH conditions. Extremophiles. 2013; 17(4): 689–696. PubMed Abstract \| Publisher Full Text\n\nMoran-Reyna A, Coker JA: Data of pH extremes effects on haloarchaea growth and transcriptomic profiles. F1000Res. 2014. Data Source" }	[ { "id": "5552", "date": "05 Aug 2014", "name": "Jesse Dillon", "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 study of responses of three well-characterized halorachaea to pH changes. Specifically, growth changes were measured and transcriptional changes were measured via cDNA microarray analysis for cells placed under different pH treatments. The authors observed some interesting differences in responses, especially in the Halorubrum strain under acidic conditions suggesting that pH responses are not universal among Halobacteriales. The finding that TFB, a putative transcription factor, is universally found under alkaline conditions may be of importance in understanding general responses to pH in haloarchaea. Overall, the manuscript is well written and the science seems sound; however, additional information in various sections of the manuscript is necessary to fully evaluate the study. Abstract:The statement \"However, of all the naturally occurring stresses faced by the haloarchaea, only one, pH, has not been previously reported on.\" is incorrect as a number of studies have reported on pH optima for described halorachaea. For example, Bowers and Wiegel published a minireview on pH and temperature responses of haloarchaea (Bowers and Wiegel, Extremophiles, 2011). I would recommend the authors qualify this statement to indicate they are the first to measure gene transcript changes in responses to pH and also include this reference and others from the Wiegel group in their article. Although the shorthand abbreviations for halophiles are well known within the halophile research community, I would define the three species using full scientific names on first usage in the abstract since F1000 has a very general audience. The statement beginning \"Our observations...\" could be improved by specifying the similarities or differences. As written the sentence is more indicative than substantive. Introduction:The extensive summary of results presented in the introduction beginning \"Broadly...\" is inappropriate in the introduction and should be deleted. Methods:More detail is needed on how the various culture media were adjusted for pH. What buffers were used? How did the authors ensure that pH did not change throughout the growth experiments? The way the description of pH manipulations is written made it seem like the optimal pH 7.2-7.5 control cultures were not included, although that is not the case based on their results. The authors should include these values in the list of treatments. Reference is made to control samples in the cDNA preparation section. Please elaborate on exactly what these controls were and how they affected the study's findings. It would be helpful if the Perl scripts described in the Microarray data processing section were made available in the supplementary information section. I may be missing something here. Reference is made to \"genes showing greater than 2-fold change\" in transcriptional level in \"at least\" 2 replicates”. However, the methods describe duplicate technical microarray replicates, so how could it be more than 2? Please clarify. Results/Discussion:Since optical densities vary among strains -- it would be useful for comparative purposes if the authors could calculate and report their growth data in terms of cells/ml as opposed to OD. What do the error bars represent in the growth curve figures? Overall, the degree of replication in the growth experiments is unclear, please clarify the statement: \"Each growth curve was measured twice and replicated once.\" Statistical tests (e.g. ANOVA) for differences among growth curves should be performed to better support statements about differences in growth rates. The section Microarray analysis and responses in R/D might be better placed in methods since it does a good job of clarifying the methodology and does not really report findings. Some interpretation of their results in the broader ecological context of pH environments encountered in nature would be warranted. For example, in the introductions, the authors note that halophiles have been isolated from sites with extremes in pH; however, their current study focuses on well characterized, \"lab rat\" haloarchaea isolated from circumneutral habitats. These strains have the advantages of characterized genomes, etc. However, in future studies it would be interesting to see if the same responses are found for organisms such as alkaliphiles isolated from more extreme pH habitats. Acknowledgment of these limitations/opportunities in the discussion would be appropriate. Presentations of Figures:Presentation of figures would be improved if the authors matched error bar colors with the line. Also, the small size of line markers and lack of contrast among some of the colors (e.g. dark shades of blue and black) makes it hard to differentiate some of the growth curves.", "responses": [ { "c_id": "943", "date": "15 Aug 2014", "name": "James Coker", "role": "Author Response F1000Research Advisory Board Member", "response": "Thank you very much for your thoughtful recommendations. We have incorporated all but one of them into the manuscript. The only suggested change not made was to report the growth curves as cells/ml. We have instead kept the OD readings as they are since this is a very standard way of reporting this type of data. We understand the issue of cell size possibly affecting the OD readings; however for the haloarchaea this is correlated with the production of gas vesicles, which are typically produced during cell densities (i.e. anaerobic conditions (OD600 >= 2.0)) well below the ones we report here." } ] }, { "id": "5557", "date": "12 Aug 2014", "name": "Mohammad Ali Amoozegar", "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 “The effects of extremes of pH on the growth and transcriptomic profiles of three haloarchaea” by Moran-Reyna and Coker describes effects of pH on growth of three genera in extremely halophilic archaea . The authors of the paper are expert in halophilic world studies. The manuscript is well-written and is quite straightforward. The language is very good. The paper was generally well structured and the results well implemented and presented.Please consider the following comments:Please add name of strains for three haloarchaea species.please add a conclusion for the manuscript.", "responses": [ { "c_id": "941", "date": "15 Aug 2014", "name": "James Coker", "role": "Author Response F1000Research Advisory Board Member", "response": "Thank you very much for you comments. We have added the names of the strains used in our studies to the materials and methods section. We have also added a conclusion section to the manuscript." } ] }, { "id": "5553", "date": "14 Aug 2014", "name": "Melanie Mormile", "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 the first study to determine the transcription response of haloarchaea when grown under extreme pH conditions. It is surprising that the pH tolerances of the select organisms have not been determined before and it was good to see that the authors did determine the pH values each of the organisms could tolerate. This data enabled the authors to determine the up-regulation and down-regulation of individual genes under the most acidic and alkaline conditions the organisms could tolerate in comparison to optimal pH conditions. The results support the authors' conclusions. The only thing that I would like to see included is a more complete discussion of how their results support their statement \"We also chose three organisms, as opposed to one, to ascertain if the response to pH was specific to the organism, the Halobacteriaceae, the Prokaryotes, or the three Domains of Life.\" in the first paragraph of the Results and Discussion. The authors touch on this a few times in the article but do not really bring it to a full conclusion.", "responses": [ { "c_id": "942", "date": "15 Aug 2014", "name": "James Coker", "role": "Author Response F1000Research Advisory Board Member", "response": "Thank you very much for you comments. We have added a bit more discussion at the end of the introduction, in the results and discussion section, and in the conclusion section about why the three organisms were selected and the nature of their transcriptomic responses." } ] } ]	1	https://f1000research.com/articles/3-168
https://f1000research.com/articles/2-222/v1	21 Oct 13	{ "type": "Case Report", "title": "Pulmonary metastases of malignant meningioma", "authors": [ "Suhail Basunaid", "Frits M.E. Franssen", "Ryan Accord", "Myrurgia Abdul Hamid", "Shekar Mahesh", "Brigitta G. Baumert", "Olaf E.M.G. Schijns", "Frits M.E. Franssen", "Ryan Accord", "Myrurgia Abdul Hamid", "Shekar Mahesh", "Brigitta G. Baumert", "Olaf E.M.G. Schijns" ], "abstract": "Meningioma accounts for approximately one-third of primary central nervous system tumors. Most meningiomas are benign, although up to one third are classified as atypical or malignant. We describe a 63-year Caucasian male presenting with pleural metastases from an intracranial meningioma. Distant metastases from meningiomas are infrequently found in clinical practice and mostly are associated with atypical or malignant meningiomas. There is no standard treatment; however surgical resection of both the primary and metastatic lesions is the safest therapy. The overall prognosis of atypical meningiomas is poor. Our patient died one week after discharge from our hospital.", "keywords": [ "A 63-year-old Caucasian man was referred to our hospital for further analysis of slowly progressing pleural effusion with a history of cough and dyspnea. The patient had a long history of epilepsy and meningioma. He was working as head in a department of administration. He was married and had two healthy kids. As a medication he took Pantoprazol", "Tegretol and Dorsolamide and had stopped smoking a long time ago." ], "content": "Case report\n\nA 63-year-old Caucasian man was referred to our hospital for further analysis of slowly progressing pleural effusion with a history of cough and dyspnea. The patient had a long history of epilepsy and meningioma. He was working as head in a department of administration. He was married and had two healthy kids. As a medication he took Pantoprazol, Tegretol and Dorsolamide and had stopped smoking a long time ago.\n\nOur patient was diagnosed with progression of a previous operated and irradiated (60 GY) atypical left parieto-occipital meningioma (WHO grade-II). Re-resection of the tumor was performed and histopathology showed a malignant meningioma (WHO grade-III). After reoperation, re-irradiation (60GY + additional 70GY) was given. One year later an asymptomatic re-recurrence was diagnosed (Figure 1A and B), for which conservative follow-up was performed without further surgical intervention.\n\nA: T1 weighted image after administration of Gadolinium based contrast fluid shows a large extra-axial enhancing lesion in the left parieto-occipital region with local mass effect. Note that the sagittal sinus seems to be invaded. B: Contrast enhanced T1 weighted image after administration of Gadolinium based contrast fluid showing a large resection cavity after the second operation and recurrent disease at the most upper margin of the resection plane with enhancing areas surrounding the sagittal sinus. C: Midthoracal CT slice in the transverse plane. Scan performed after i.v. administration of iodine contrast. The lesion is easily distinguished at the left ventral thoracal intrapleural space, slightly enhanced suggesting solid tissue. Some pleural fluid is also present.\n\nA few months later the patient was hospitalized with dyspnea, fatigue, productive cough and anorexia. Multiple pleural masses were detected at a chest computer-tomography (CT) scan. Histopathology was consistent with malignant meningioma (WHO grade-III, Figure 2A), there was a high expression in the EMA staining, also in the AE1/AE3 staining (Figure 2B). The CD 45 and CD 68 were positive and MIB-1 showed high proliferation. Palliative chemotherapy was offered but refused by the patient. The patient is died one week after discharge from the hospital as a result of voluntary euthanasia as was the will of the patient (valid written declaration).\n\nA: A specimen (HE-stained, 40x) showing histological resemblance between the intrathoracal lesion and the intracranial meningioma. B: Specimen (focal plus and focal weak) showed high expression in the EMA-staining and also in the AE1/AE3-staining.\n\n\nDiscussion\n\nPulmonary and pleural metastases from an intracranial meningioma are very rare. Distant metastases from meningiomas are infrequently found in clinical practice and mostly associated with atypical or malignant meningiomas. Meningiomas mainly recur loco-regional or adjacent to the radiation treatment fields1,2. There are only isolated case reports regarding pulmonary metastases from meningioma. Most lung metastases were incidentally detected by chest radiography or by CT-scans, because metastatic lesions are usually asymptomatic. The presence of pulmonary metastases appears to negatively affect survival in patients with recurrent meningioma3,4.\n\nRegarding the relationship between the intracranial location and invasion of the sagittal sinus of the tumor and the pleural metastases the route of dissemination is most probably the central venous route to heart and lungs5. In previous case reports the lung was the most common extracranial metastatic site for intracranial meningioma6.\n\nIn this case, histopathologic findings of the primary tumor revealed hypercellularity, wide necrosis, and brain invasion into the normal brain parenchyma. Pathology of the lesion from the left thoracic wall was consistent with malignant meningioma.\n\nOther case studies described that treatment of pulmonary metastasis of malignant meningioma consisted of surgical resection for both the primary or metastatic lesions1,5,7.\n\nPostoperative conventional radiation therapy has been recommended for prevention of local recurrence, especially when resection is subtotal.\n\n\nInformed consent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the next of kin.", "appendix": "Author contributions\n\n\n\nFrits Franssen: supervisor of the respiratory department where patient was admitted. Ryan Accord: the surgeon who did the VATS. Myurgia A. Hamid: histological figures. Shekar Mahesh: radiological figures. Brigitta G. Baumert and Olaf E.M.G. Schijns: literature search and data collection.\n\n\nCompeting interests\n\n\n\nNo competing interests have been disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nAt the completion of this Case study, I am very thankful to all contributors, who were very helpful and without their support this case study would have never come into its present form.\n\n\nReferences\n\nFigueroa BE, Quint DJ, Mckeever PE: Extracranial metastatic meningioma. Br J Radiol. 1999; 72(857): 513–516. PubMed Abstract\n\nStoller JK, Kavuru M, Mehta AC: Intracranial meningioma metastatic to the lung. Cleve Clin J Med. 1987; 54(6): 521–526. PubMed Abstract \| Publisher Full Text\n\nAlexandru D, Glantz MJ, Kim L: Pulmonary metastases in patients with recurrent, treatment-resistant meningioma: prognosis and identification by 111Indium-octreotide imaging. Cancer. 2011; 117(19): 4506–11. PubMed Abstract \| Publisher Full Text\n\nEnam SA, Abdulrauf S, Mehta B: Metastasis in meningioma. Acta Neurochir (Wien). 1996; 138(10): 1172–7. PubMed Abstract \| Publisher Full Text\n\nNakano M, Tanaka T, Nakamura A: Multiple Pulmonary Metastases following total removal of a bilateral parasagittal meningioma with complete occlusion of the superior sagittal sinus: Report of a Case. Hindawi Publishing Corporation Case Rep Neurol Med. 2012; 2012. : 7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStrang RR, Tovi D, Nordenstam H: Meningioma with intracerebral, cerebellar and visceral metastases. J Neurosurg. 1964; 21(12): 1098–1102. PubMed Abstract \| Publisher Full Text\n\nKanzaki R, Higashiyama M, Fujiwara A: Surgical resection of pulmonary metastases from meningioma: report of a case. Surg Today. 2011; 41(7): 955–8. PubMed Abstract \| Publisher Full Text" }	[ { "id": "2143", "date": "29 Oct 2013", "name": "Daniel Vorobiof", "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 case report documents an unusual spread of a malignant meningioma. It doesn't add any further information and doesn't contribute any new knowledge, neither diagnostic nor therapeutic to the current available medical literature.", "responses": [ { "c_id": "854", "date": "06 Jun 2014", "name": "Suhail Basunaid", "role": "Reader Comment", "response": "Dear Sir,I am not sure if you consider this an unusual spread of malignant meningioma. There were at least 3 years between the time of discovering it (by accident after a fall from his bicycle) and the pulmonary involvement. I can understand that this case report might not be a unique case but I found myself obliged to report it. There is not enough knowledge in the literature regarding the therapeutic options for malignant meningioma with distance metastases. There was a declaration of euthanasia by our patient but things went too quickly and our patient died shortly after discharge before being able to make an accurate decision." } ] }, { "id": "2861", "date": "16 Jan 2014", "name": "Matthias Simon", "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\nBasunaid et al. describe a patient with a recurrent meningioma who ultimately developed pleural metastases. The primary tumor was assigned to WHO grade II, and the recurrent tumor to grade III. While this paper by itself does not provide truly novel information, in my view such rare patients should be reported in the literature. This and similar reports will help to build a database which may at some point help to improve our understanding and management of rare conditions. The report may benefit from a more comprehensive and detailed (tabular?) review of the literature.", "responses": [ { "c_id": "881", "date": "27 Jun 2014", "name": "Suhail Basunaid", "role": "Author Response", "response": "I would like to thank Dr. Simon for his time and valuable comment. I wish to leave my current version as it is." } ] }, { "id": "5253", "date": "27 Jun 2014", "name": "Christine Marosi", "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 fully agree with the reviewer from Bonn, Matthias Simon. Meningiomas grade III are fortunately rare and they do metastasize through vascular pathways. Of course a register of such cases would be a valuable tool to get an idea of the prevalence of such events. The case reported by Basunaid is a tragic one as the patient asked for euthanasia one week after dismissal from hospital. This shows that he could not be offered any therapeutic option able to provide enough solace for staying alive.", "responses": [ { "c_id": "882", "date": "27 Jun 2014", "name": "Suhail Basunaid", "role": "Author Response", "response": "I would like to thank Dr. Marosi for her time spent on reviewing this case report, and the valuable comments given. I would like to leave my current version as it is. I completely agree regarding the facts about the rarity of the case and the termination by the euthanasia declaration." } ] }, { "id": "5747", "date": "21 Aug 2014", "name": "Antonio Santacroce", "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\nBasunaid et al. report on a patient harboring a recurrent meningioma who developed pleural metastases. The primary tumour is reported to be an atypical meningioma WHO Gr II, which upon recurrence shows a malignant histology WHO Gr. III.I share the opinion of Drs Marosi and Simon about the importance of continuing reporting such clinical cases due to the rarity of such tumour entities, even though, as reported by Dr Vorobiof \"It doesn't add any further information and doesn't contribute any new knowledge, neither diagnostic nor therapeutic to the current available medical literature\" There are some points which should be clarified. To quote the authors:\"Our patient was diagnosed with progression of a previous operated and irradiated (60 GY) atypical left parieto-occipital meningioma (WHO grade-II). Re-resection of the tumor was performed and histopathology showed a malignant meningioma (WHO grade- III). After reoperation, re-irradiation (60GY + additional 70GY) was given.\" Which radiation technique was used? It is not clear whether the radiation dose applied upon recurrence was only 60Gy with the boost, and if the “additional” 70 Gy is the cumulative dosage applied including the boost, or just the boost itself. I would also recommend verifying in the literature whether radiation therapy has been offered for histologically confirmed metastatic extracranial lesions.", "responses": [] } ]	1	https://f1000research.com/articles/2-222
https://f1000research.com/articles/3-126/v1	11 Jun 14	{ "type": "Research Article", "title": "The assessment of CD146-based cell sorting and telomere length analysis for establishing the identity of mesenchymal stem cells in human umbilical cord", "authors": [ "Dimitrios Kouroupis", "Sarah M. Churchman", "Dennis McGonagle", "Elena A. Jones", "Dimitrios Kouroupis", "Sarah M. Churchman", "Dennis McGonagle" ], "abstract": "Adult stem cells are characterised by longer telomeres compared to mature cells from the same tissue. In this study, candidate CD146+ umbilical cord (UC) mesenchymal stem cells (MSCs) were purified by cell sorting from UC tissue digests and their telomere lengths were measured in comparison to donor-matched CD146-negative fraction.UC tissue fragments were enzymatically treated with collagenase and the cells were used for cell sorting, colony-forming fibroblast (CFU-F) assay or for long-term MSC cultivation. Telomere lengths were measured by qPCR in both culture-expanded MSCs and candidate native UC MSCs. Immunohistochemistry was undertaken to study the topography of CD146+ cells.Culture-expanded UC MSCs had a stable expression of CD73, CD90 and CD105, whereas CD146 declined in later passages which correlated with the shortening of telomeres in the same cultures. In three out of four donors, telomeres in candidate native UC MSCs (CD45-CD235α-CD31-CD146+) were longer compared to donor-matched CD146- population (CD45-CD235α-CD31-CD146-). The frequency of CD45-CD235α-CD31-CD146+ cells measured by flow cytometry was ~1000-fold above that of donor-matched CFU-Fs (means 10.4% and 0.01%, respectively). CD146+ cells were also abundant in situ having a broad topography including high levels of positivity in muscle areas in addition to vessels.", "keywords": [ "Perinatal mesenchymal stem cell (MSC) sources are attracting increasing attention as an alternative to ‘gold standard’ bone marrow (BM) MSCs. Owing to its universal accessibility", "umbilical cord (UC) tissue represents an attractive source of MSCs for cell therapy1. However", "the potency of standard culture-expanded UC MSCs", "especially towards osteoblasts", "chondrocytes and adipocytes", "is lower compared to BM MSCs2", "3. This can be explained by the fact that conventional UC MSC cultures arise from diverse clonal populations that possess varying degrees of self-renewal leading to mixed cultures that gradually lose their MSC properties4. The therapeutic advantages of minimally-expanded UC MSCs would therefore include a better preservation of their native functionality as well as rapid manufacture and reduced cost2. However", "no agreement yet exists on the native phenotype of UC MSCs5–7", "which is an essential pre-requisite for their isolation and minimal expansion." ], "content": "Introduction\n\nPerinatal mesenchymal stem cell (MSC) sources are attracting increasing attention as an alternative to ‘gold standard’ bone marrow (BM) MSCs. Owing to its universal accessibility, umbilical cord (UC) tissue represents an attractive source of MSCs for cell therapy1. However, the potency of standard culture-expanded UC MSCs, especially towards osteoblasts, chondrocytes and adipocytes, is lower compared to BM MSCs2,3. This can be explained by the fact that conventional UC MSC cultures arise from diverse clonal populations that possess varying degrees of self-renewal leading to mixed cultures that gradually lose their MSC properties4. The therapeutic advantages of minimally-expanded UC MSCs would therefore include a better preservation of their native functionality as well as rapid manufacture and reduced cost2. However, no agreement yet exists on the native phenotype of UC MSCs5–7, which is an essential pre-requisite for their isolation and minimal expansion.\n\nProspective isolation of UC MSCs has been attempted after enzymatic digestion followed by selection for specific cell surface markers using fluorescence-activated cell sorting (FACS) or immunomagnetic bead separation techniques. Previous studies showed that CD45-based negative depletion5, neural ganglioside (GD2)-7 and CD146-based positive selection6–11 could be used to select for clonogenic and multipotential MSC fractions. Although CD146 is expressed on pericytes, which have been proposed as a reservoir of tissue specific progenitors and MSCs12,13, it is also present on the surface of CD31+ UC endothelial cells (ECs)2,6,9, and therefore cannot be used alone to achieve high-purity UC MSC preparations.\n\nWhilst the majority of studies in this field used combinations of surface markers to isolate putative MSCs from different tissues, followed by their long-term clonal expansion and analysis of multipotentiality10,14, another alternative and faster approach could involve measuring the telomere length of freshly-sorted candidate MSC fractions. As immature cells, MSCs by definition should have longer telomeres compared to mature cell types. This idea was first proposed by Flores et al. (2008) who showed that stem cell compartments from hair follicles, intestine, testis, cornea and brain are enriched with cells with the longest telomeres15.\n\nThe aims of this study were to confirm that CD146 could be a good MSC marker in the UC and to purify candidate MSCs from UC tissue digests based on the non-haemopoietic (CD45-CD235α-), non-EC (CD31-), CD146+ phenotype. We measured telomere lengths in CD146+ and donor-matched control CD146- populations using quantitative real-time PCR (qPCR) and examined the topography of CD146+ cells in situ using immunohistochemistry.\n\n\nMaterials and methods\n\nUC tissue was collected from the UCs of consenting full-term caesarean section patients (n=8). After delivery, UCs were immediately stored in Dulbecco's phosphate buffered saline (DPBS, #14190-250, Invitrogen, Renfrew, UK) at 4°C. All samples were obtained after written informed consent and the protocols were approved by National Research Ethics Committee (07/Q1205/27). Human primary skin fibroblasts were obtained from Lonza and ATCC (Lonza, Cambridge, UK and ATCC, Middlesex, UK).\n\nThe whole UC tissue was mechanically dissected in small pieces (~0.2±0.1 g) and washed repeatedly with DPBS followed by enzymatic digestion using 600 U/ml collagenase I (#07902, Stem Cell Technologies, Grenoble, France)2 for 6 hours. Released cells were resuspended in 1:50 v/v DPBS, filtered through a 70 μm cell strainer (#352350, BD Biosciences, Oxford, UK), centrifuged (650 g) and counted.\n\nFor primary MSC culture, UC tissue digests were seeded in 6-well tissue culture plates (#3516, Corning, Amsterdam, The Netherlands) in non-haematopoietic (NH) expansion medium (#130-091-680, Miltenyi Biotec, Bisley, UK) at a density of 5×104 cells/well. After observing 80% cell confluency [denoted passage (p) 0], adherent cells were trypsinised using 0.5% trypsin/EDTA (#15400-054, Invitrogen) and re-seeded at 5×105/25 cm2 flask for further passaging which was performed as 1:1 splits until approximately p17. Cultures were fed twice weekly with half media changes. The number and viability of the cells at each passage were evaluated using Trypan blue staining. For primary EC culture, 5×104 cells were seeded into 6-well tissue culture plates coated with 2 μg/cm2 fibronectin (#354008, BD Biosciences) in endothelial basal medium (EBM2, #CC-3162, Lonza). On day 2, adherent cells were washed with DPBS and subsequently fed three times per week with half media changes. ECs were passaged using 1:1 splits until p4.\n\nFor all cultures, the number of population doublings (PDs) between passages (starting from initial passage; p0) was calculated according to the following equation: PD= log 2 (Nt/Ni), where Ni and Nt are the initial and terminal cell counts, respectively. PDs before p0 were calculated based on colony-forming unit-fibroblast (CFU-F) potential of cells seeded and the number of cells at p0 according to the equation PD= log 2 (N cells at p0/N seeded CFU-Fs). The CFU-F assay was performed in triplicate at the cell seeding density of 5×104 cells/well and individual colonies (>50 cells) were counted on day 14 after staining with 1% w/v Crystal Violet (#V5265, Sigma, Hertfordshire, UK).\n\nRNA extraction from culture expanded UC MSCs (p4) and fibroblasts was performed using RNA/DNA/Protein purification kit (#23500, Norgen, Ontario, Canada) according to the manufacturer’s instructions and RNA yield was quantified by using NanoDrop 2000 spectrophotometer (Thermo, Essex, UK). Single strand cDNA was synthesised using high-capacity cDNA reverse transcription kit (#4368814, Applied Biosystems, Warrington, UK).\n\nA custom designed 48 gene Taqman low density array (TLDA, Applied Biosystems) contained mesenchymal and endothelial lineage-related transcripts and novel surface receptors that could be used to segregate MSCs from fibroblasts based on previously published data16,17. To perform TLDA, 200 ng cDNA were used per port. The results were obtained using an ABI PRISM 7900HT SDS (Applied Biosystems). Normalisation of transcript levels relative to reference gene HPRT was performed using the formula: 2-ΔCt, ΔCt=Ct value (selected transcript) - Ct value (HPRT).\n\nPhenotypic characterisation was performed on culture expanded MSCs at different passages and on cultured ECs at p4 using: CD31-FITC (#MCA1738F), CD105-PE (#MCA1557PE), CD90-PE (#MCA90PE) (all from Serotec, Kidlington, UK), CD73-PE (#550257), CD146-PE (#550315) (both from BD Pharmingen, Oxford, UK), and CD271-PE (#130-091-885, Miltenyi Biotec). The isotype controls were IgG1-FITC (#550616, BD Pharmingen) and IgG1-PE (#MCA928PE, Serotec). A total of 2×105 cells was stained with 5 μl FITC- or PE-conjugated antibodies, and dead cells were excluded using 2 μg/ml propidium iodide (PI, #P1304MP, Invitrogen). Cells were acquired using FACScan equipped with CellQuest software version 3.1 (BD Biosciences) and the proportions of the different fractions were calculated as a percentage of total live cells.\n\nCell sorting was performed using a MoFlo cell sorter equipped with SUMMIT software (Beckman Coulter, Buckinghamshire, UK). Following collagenase digestion of UC tissue, 2×106 cells were split into two tubes. One tube was stained with 5 μl of neat CD45-FITC (#F0861), CD235α-FITC (#F0870) (both from DAKO, Cambridge, UK), CD146-PE (BD Pharmingen) and CD31-APC (#130-092-652, Miltenyi Biotec), whereas the other was stained with 2.5 μl of neat isotype controls alone. After incubation with relevant antibodies and washes, 2 μg/ml 7-aminoactinomycin D (7-AAD) (#A1310, Invitrogen) was added to exclude dead cells before sorting into four fractions: haemopoietic cell fraction (HC), CD45+CD235α+CD31-; EC fraction, CD45-CD235α-CD31+; double-negative candidate MSC fraction, CD45-CD235α-CD31-CD146+ and non-MSC fraction, CD45-CD235α-CD31-CD146-. The latter two subsets were processed for telomere length analysis.\n\nQIAamp DNA Mini kit (#51306, Qiagen, Crawley, UK) was used for gDNA extraction from cultured MSCs and freshly-sorted CD146+ and CD146- subsets. Telomere length measurement by SYBR Green qPCR (#4309155, Invitrogen) involved determining the relative ratio of telomere repeat copy number (T) to a single copy gene (36B4) copy number (S): T/S ratio, as previously described18,19. Samples were run in triplicate using 20 ng gDNA.\n\nImmunohistochemistry was used to characterise UC tissue architecture and investigate CD146 cell topography in situ to ascertain whether it exhibited the proposed MSC pericyte distribution. Whole UC tissue cross sections were embedded in OCT mounting media (#361603E, VWR, Leicestershire, UK), snap frozen in liquid nitrogen and stored at -80°C. Cryostat sections (6 μm) were mounted on superfrost slides (#48311-703, VWR) and dried overnight at 37°C. Immunohistochemistry was performed using DAKO REAL detection system (#K4065, DAKO) according to the manufacturer’s instructions. Primary antibodies included: CD31 (#CBL468, working concentration 1:10), CD146 (#MAB16985, working concentration 1:50) (both from Chemicon, Watford, UK), CD34 (#M716501, working concentration 1:100), CD271 (#M3507, working concentration 1:20) (both from DAKO), CD90 (#MCA90, working concentration 1:100) and alpha smooth muscle actin (αSMA, #MCA5781GA, working concentration 1:20) (both from Serotec). Antibody binding was visualized using DAKO REAL DAB+ chromogen (#K3468, DAKO) and slides were counterstained by mounting in Harris haematoxylin (#HHS128, Sigma). Slides were mounted using DPX (#317616, Sigma) and images were captured using CAMEDIA C-7070 camera (Olympus, Tokyo, Japan).\n\nThe software used for analysis and statistics was GraphPad (GraphPad software, La Jolla, USA). The gene expression results were analysed with Mann-Whitney test for unpaired samples (p<0.05: high significance, 95% confidence interval). The cell sorting and CFU-F results were analysed by Kruskal-Wallis test with Dunn’s correction.\n\n\nResults\n\nWe initially aimed to confirm the validity of CD146 as a candidate marker of UC MSCs using cultures established following standard MSC protocols2. Mesenchymal tri-potentiality of these cultures was demonstrated in our previous study2. We studied the expression of 45 MSC- and fibroblast-related transcripts by TLDA in these cultures and compared their expression levels to those of negative control skin fibroblasts. Eighteen transcripts were selected (Table 1) as they showed interesting differences between UC MSCs and fibroblasts. Moreover, five transcripts showed more than two-fold higher expression in UC MSCs. These included NANOG (a pluripotency marker), whose expression in UC cultures suggested their greater level of immaturity, as well as NGFR/CD271 and MCAM/CD146, the markers of native BM MSCs20,21. Fibroblasts showed stronger expression of eight transcripts (Table 1).\n\nNormalisation of transcript levels relative to reference gene HPRT was performed using the formula: 2-ΔCt, ΔCt=Ct value (selected transcript) - Ct value (HPRT).\n\nHaving shown NGFR and MCAM transcript expression in culture-expanded UC MSCs, their surface protein expression at p4 and subsequent passages was next investigated. Phenotypic characterization was performed at four different passages (representing approximately 17, 19, 21, 25 PDs). MSC-specific markers CD90, CD73 and CD10522 were uniformly positive showing a stable expression profile throughout the expansion, whereas the absence of EC marker CD31 showed no contamination with ECs (Figure 1A).\n\nA - Surface expression levels of MSC markers (n=3 donors, y error bars indicate SD). B - Donor variation of CD146 and CD90 markers in early (<20 PDs) and late (>20 PDs) cultures (M1: marker expression). Bottom panels - telomere T/S ratios were directly correlated to the expression levels of CD146 and not CD90 during culture-expansion.\n\nDespite the expression of NGFR transcript, CD271 surface protein expression was absent in culture-expanded MSCs at all tested time-points (Figure 1A). CD146 expression declined gradually and correlated with telomere loss in the two cultures tested (Figures 1A and 1B). This effect was not evident for CD90 which remained stable (Figure 1B). When comparing the phenotypic profile of UC MSCs with ECs only CD31 revealed high specificity for ECs; CD73, CD105, CD146 were expressed on both UC MSCs and ECs (data not shown).\n\nIn combination with the molecular profile of UC MSCs and the previously-published literature6,8–11, these data indicated that uncultured UC CD146+ cells could indeed possess longer telomeres than the remaining CD146- cells.\n\nThe cell sorting strategy for these experiments is described in Figure 2A; live cells (R1), containing distinct populations of ECs (R2) and HCs (R3) were clearly observed. Double negative cells (R4) were further subdivided into CD146+ (R5, candidate UC MSC) and CD146- (R6, non-MSC) subsets and sorted into separate RNA lysis buffers. In general, HCs were most abundant (mean 48% of total live cells, n=8). CD31+ ECs represented a mean of 1.4% of live cells.\n\nA - Cell sorting strategy: live cells (upper right graph, R1) were identified by 7AAD exclusion method. Three distinct populations (bottom right graph, R2/ECs, R3/HCs, and R4/double=negative) were evident. Gating on double-negative subset (R4) revealed two subsets (bottom left graph), CD146+ (R5/candidate MSCs) and CD146- (R6/non MSCs). Cells confined to regions R5 and R6 were sorted and processed for telomere length analysis. B - The percentage of CD146+ and CD146- populations compared to the percentage of CFU-Fs (n=4 donors, y error bars indicate SD, p<0.05, **p<0.001). C - Telomere T/S ratios in sorted subsets (n=4 donors).\n\nWithin the double-negative cells, the CD146- population was predominant over the CD146+ population (mean 81% and 17% respectively, p<0.05; Figure 2A). When these frequencies were re-calculated in relation to total live cells, the CD146+ and CD146- fractions represented a mean of 10.4% and 37.7%, respectively (Figure 2B). This was significantly higher than the frequency of CFU-Fs (as a percentage of total live cells) in the same UC digests (mean 0.01%, Figure 2B). When the telomere length of both sorted subsets (CD146+ and CD146-) were tested, the CD146+ subset exhibited higher telomere lengths compared to the CD146- subset in three out of four donors (Figure 2C), hence the observed differences failed to reach statistical significance.\n\nThe high frequency of CD45-CD31-CD235α-CD146+ cells compared to the CFU-F frequency and the lack of consistent and significant enrichment for cells with long telomeres indicated that the CD146+ UC fraction most likely contained non-haemopoietic cells with varying degrees of maturity.\n\nInitial haematoxylin staining (Figure 3A) revealed the basic structure of the UC tissue where the UC vein and artery could be seen. This was surrounded by the single thickness endothelial area (EA) and highly organised muscular perivascular area (PA). The Wharton’s jelly (WJ) matrix could also be seen spanning intra-muscular areas.\n\nA – Images show endothelial area (EA, indicated by arrows), perivascular area (PA, multiple layers of muscle fibres) and Wharton’s jelly area. (B) Expression of MSC and EC markers in UC tissue. (C) Staining of MSC and EC markers in UC tissue (representative donor and cross sections, original magnification ×200; (+) symbol indicates the expression of a marker and (-) symbol indicates the absence of expression of a marker).\n\nImmunohistochemistry was next used to investigate marker expression in situ including semi-quantitative assessment in different anatomical areas (Figures 3B and 3C). Three optical microscope fields (×200) were evaluated per anatomical area of UC tissue. CD271 was present at low levels in the WJ area only. EC-specific CD31 and CD34 were present in the EA of the vessels. The intracellular marker of smooth muscle cells; αSMA23 was expressed and had its highest positivity in the PA surrounding vessels. CD90 and CD146 were highly expressed in most UC compartments, including the PA; however CD146 but not CD90 showed positivity for the EA area. This was consistent with previously published literature9. Representative photomicrographs are shown in Figure 3C.\n\nOverall, our immunohistochemistry results revealed the expected topography of CD31+ and CD34+ ECs, the expression of CD271 in the WJ area, and the preferential topography of CD90 and αSMA in PA. Consistent with previous findings9, CD146 was expressed in WJ, PA and EA, with the highest proportion of cells present in the PA. Wide distribution of CD146+ cells in all anatomical areas of UC tissue was consistent with the high frequency of CD45-CD31-CD235α-CD146+ cells evident by flow cytometry. This indicated that UC MSC isolation alone, even after the removal of CD31+ ECs, was not sufficient to purify native MSCs from the UC tissue.\n\n\n\n\nDiscussion\n\nThe present study assessed the possibility of using CD146-based cell sorting and telomere length analysis for establishing the identity of mesenchymal stem cells in human UC. Previous studies have shown that the UC CD146+ subset contained MSCs able to differentiate into osteoblasts, chondrocytes and adipocytes8,9. Other studies have shown that ‘true’ immature stem cells have much longer telomeres compared to the remaining mature cells15,24. Here we tested whether CD146+ UC cells, that were depleted of contaminating haemopoietic and endothelial cells, had longer telomeres compared to the corresponding CD146- population.\n\nInitially we performed gene expression analysis, flow cytometry and telomere length measurements on culture-expanded UC MSCs during their extensive passaging. The mesenchymal tripotentiality of these cultures has been demonstrated in our previous study2. In this study, gene expression analysis revealed that MCAM/CD146 was expressed at higher levels (>3-fold) in UC MSCs compared to fibroblasts, confirming its potential specificity for MSCs. Whereas common MSC-specific markers CD73, CD105, CD9022 displayed stable expression throughout passaging, CD146 expression declined with the increased number of cell divisions at later passages consistent with a loss of multipotential progenitors4. The CD271 surface marker was not expressed at any stage during UC MSC cultivation, consistent with the loss observed in BM MSCs16,25,26. Although CD271 is expressed on uncultured BM MSCs, it was not selective for UC blood MSCs21,27. Therefore, CD271-based MSC isolation from UC tissue was not pursued further. However, the decrease in CD146 during expansion correlated with telomere erosion in the same cultures, supporting the idea that CD146 could mark the most immature cells in vivo. Based on the present findings and previous data9,11, CD146 was selected as a candidate positive marker for sorting native MSCs from UC digests. CD45/CD235α and CD31 were used to exclude native HCs and ECs from the analysis, respectively2.\n\nSorted CD146+ and CD146- subsets were next compared with respect to their telomere length. Several previous studies provided initial evidence that UC MSCs express telomerase continuously and hence were able to retain long telomeres28,29. In UC blood MSCs, a distinct SSEA-4+CD105+MSCA-1-CD90- cell population was shown to have longer telomeres than the SSEA-4-CD105+MSCA-1+CD90+ subset30. Although evident in three out of four cultures, in our study we observed the lack of significant difference in telomere length between CD146+ and CD146- subsets potentially indicating contamination of MSCs with more mature cells in the CD146+ subset. The high frequency of CD146+ cells (10.4%) was in stark contrast to the CFU-F frequency (~0.01%), which was consistent with previous findings5,31. Furthermore, the immunohistochemistry data confirmed the broad reactivity of CD146 with UC tissue anatomical areas9. Altogether, these data provided the first indication that the CD146+ population was unlikely to represent pure native UC MSCs.\n\nAlternatively, a lack of significant difference in telomere lengths could be explained by the limited sensitivity of qPCR assay to measure telomere lengths32. Telomere length assessment in the majority of studies using cultured MSC is based on terminal restriction fragment (TRF) analysis33–37, which has the disadvantage of producing a smear of bands rather than a discrete band/point affecting accurate quantifications. Additionally, the large number of cells required for such analysis precluded its use on sorted UC cells. On the other hand, studies investigating the correlations between telomere lengths derived by qPCR and TRF indicated good levels of correlation (r>0.823)20,38, although qPCR was shown to be limited in its ability to measure the longest telomeres32. This potential technical limitation of qPCR could have affected the accuracy of the telomere analysis in the present study.\n\nOur findings could be further compounded by the large donor-to-donor variation, which was evident in culture expanded UC MSCs (~10% between donor 1 and 2, Figure 1B), but also with sorted subsets (Figure 2C). One recent study has demonstrated that telomere lengths could be heritable, with the parental age at conception being a factor affecting offsprings’ telomere length in leukocytes (LTL)38. Although the mode of LTL inheritance has been suggested to be X-linked39, another study shows a paternal mode of heritability40. The parental ages of the UC donors were not known/recorded in this study but would be an interesting subject for the future work. Previous studies have suggested that the magnitude of inter-individual variation in telomere lengths could exceed the variation between cell types within the same individual41–43, therefore telomere length measurement approaches should be further refined before they can be used as a tool to identify MSCs in vivo.\n\nTo conclude, the broad availability of UC tissue makes it amenable to be used in cell therapy and regenerative medicine interventions. To this end, the present study showed that in the UC, CD146+ cells were too numerous to be selective for pure native MSCs and were likely to contain more mature cells. Therefore, additional markers would be needed to isolate MSCs from UC tissue.\n\n\nData availability\n\nF1000Research: Dataset 1. Data for CD146-based cell sorting and telomere length in umbilical cord, 10.5256/f1000research.4260.d2870044\n\n\nConsent\n\nWritten informed consent has been obtained from full-term caesarean section patients for the use of clinical samples in the present study.", "appendix": "Author contributions\n\n\n\nDK, SMC, DM, EJ conceived the study. DK performed the experimental work. All authors were involved in manuscript preparation and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nSC is supported by NIHR-Leeds Musculoskeletal Biomedical Research Unit (LMBRU). EJ and DM hold funding from Wellcome Trust/EPSRC through WELMEC, a Centre of Excellence in Medical Engineering, under grant number WT088908/Z/09/Z and are additionally supported by LMBRU.\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 Mrs. Anne English for her valuable technical contribution in histology and flow cytometry in this project.\n\n\nReferences\n\nTroyer DL, Weiss ML: Wharton’s jelly-derived cells are a primitive stromal cell population. Stem Cells. 2008; 26(3): 591–599. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKouroupis D, Churchman SM, English A, et al.: Assessment of umbilical cord tissue as a source of mesenchymal stem cell/endothelial cell mixtures for bone regeneration. Regen Med. 2013; 8(5): 569–581. PubMed Abstract \| Publisher Full Text\n\nBosch J, Houben AP, Radke TF, et al.: Distinct differentiation potential of “MSC” derived from cord blood and umbilical cord: are cord-derived cells true mesenchymal stromal cells? Stem Cells Dev. 2012; 21(11): 1977–1988. PubMed Abstract \| Publisher Full Text\n\nSarugaser R, Hanoun L, Keating A, et al.: Human mesenchymal stem cells self-renew and differentiate according to a deterministic hierarchy. PLoS One. 2009; 4(8): e6498. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSarugaser R, Lickorish D, Baksh D, et al.: Human umbilical cord perivascular (HUCPV) cells: a source of mesenchymal progenitors. Stem Cells. 2005; 23(2): 220–229. PubMed Abstract \| Publisher Full Text\n\nBaksh D, Yao R, Tuan RS: Comparison of proliferative and multilineage differentiation potential of human mesenchymal stem cells derived from umbilical cord and bone marrow. Stem Cells. 2007; 25(6): 1384–1392. PubMed Abstract \| Publisher Full Text\n\nXu J, Liao W, Gu D, et al.: Neural ganglioside GD2 identifies a subpopulation of mesenchymal stem cells in umbilical cord. Cell Physiol Biochem. 2009; 23(4–6): 415–424. PubMed Abstract \| Publisher Full Text\n\nTsang WP, Shu Y, Kwok PL, et al.: CD146+ human umbilical cord perivascular cells maintain stemness under hypoxia and as a cell source for skeletal regeneration. PLoS One. 2013; 8(10): e76153. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchugar RC, Chirieleison SM, Wescoe KE, et al.: High harvest yield, high expansion, and phenotype stability of CD146 mesenchymal stromal cells from whole primitive human umbilical cord tissue. J Biomed Biotechnol. 2009; 2009: 789526. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTormin A, Li O, Brune JC, et al.: CD146 expression on primary nonhematopoietic bone marrow stem cells is correlated with in situ localization. Blood. 2011; 117(19): 5067–5077. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCovas DT, Panepucci RA, Fontes AM, et al.: Multipotent mesenchymal stromal cells obtained from diverse human tissues share functional properties and gene-expression profile with CD146+ perivascular cells and fibroblasts. Exp Hematol. 2008; 36(5): 642–654. PubMed Abstract \| Publisher Full Text\n\nCrisan M, Yap S, Casteilla L, et al.: A perivascular origin for mesenchymal stem cells in multiple human organs. Cell Stem Cell. 2008; 3(3): 301–313. PubMed Abstract \| Publisher Full Text\n\nBianco P, Robey PG, Simmons PJ: Mesenchymal stem cells: revisiting history, concepts, and assays. Cell Stem Cell. 2008; 2(4): 313–319. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZimmerlin L, Donnenberg VS, Pfeifer ME, et al.: Stromal vascular progenitors in adult human adipose tissue. Cytometry A. 2010; 77(1): 22–30. PubMed Abstract \| Publisher Full Text\n\nFlores I, Canela A, Vera E, et al.: The longest telomeres: a general signature of adult stem cell compartments. Genes Dev. 2008; 22(5): 654–667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChurchman SM, Ponchel F, Boxall SA, et al.: Transcriptional profile of native CD271+ multipotential stromal cells: Evidence for multiple fates, with prominent osteogenic and Wnt pathway signaling activity. Arthritis Rheum. 2012; 64(8): 2632–2643. PubMed Abstract \| Publisher Full Text\n\nIshii M, Koike C, Igarashi A, et al.: Molecular markers distinguish bone marrow mesenchymal stem cells from fibroblasts. Biochem Biophys Res Commun. 2005; 332(1): 297–303. PubMed Abstract \| Publisher Full Text\n\nJones E, English A, Churchman SM, et al.: Large-scale extraction and characterization of CD271+ multipotential stromal cells from trabecular bone in health and osteoarthritis: implications for bone regeneration strategies based on uncultured or minimally cultured multipotential stromal cells. Arthritis Rheum. 2010; 62(7): 1944–54. PubMed Abstract \| Publisher Full Text\n\nCawthon RM: Telomere measurement by quantitative PCR. Nucleic Acids Res. 2002; 30(10): e47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBianco P, Sacchetti B, Riminucci M: Osteoprogenitors and the hematopoietic microenvironment. Best Pract Res Clin Haematol. 2011; 24(1): 37–47. PubMed Abstract \| Publisher Full Text\n\nWatson JT, Foo T, Wu J, et al.: CD271 as a marker for mesenchymal stem cells in bone marrow versus umbilical cord blood. Cells Tissues Organs. 2013; 197(6): 496–504. PubMed Abstract \| Publisher Full Text\n\nDominici M, Le Blanc K, Mueller I, et al.: Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy. 2006; 8(4): 315–317. PubMed Abstract \| Publisher Full Text\n\nHinz B, Celetta G, Tomasek JJ, et al.: Alpha-smooth muscle actin expression upregulates fibroblast contractile activity. Mol Biol Cell. 2001; 12(9): 2730–2741. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWagner W, Bork S, Horn P, et al.: Aging and replicative senescence have related effects on human stem and progenitor cells. PLoS One. 2009; 4(6): e5846. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuirici N, Soligo D, Bossolasco P, et al.: Isolation of bone marrow mesenchymal stem cells by anti-nerve growth factor receptor antibodies. Exp Hematol. 2002; 30(7): 783–791. PubMed Abstract\n\nJones EA, Kinsey SE, English A, et al.: Isolation and characterization of bone marrow multipotential mesenchymal progenitor cells. Arthritis Rheum. 2002; 46(12): 3349–3360. PubMed Abstract \| Publisher Full Text\n\nAttar A, Ghalyanchi Langeroudi A, Vassaghi A, et al.: Role of CD271 enrichment in the isolation of mesenchymal stromal cells from umbilical cord blood. Cell Biol Int. 2013; 37(9): 1010–1015. PubMed Abstract \| Publisher Full Text\n\nJo C, Kim OS, Park EY, et al.: Fetal mesenchymal stem cells derived from human umbilical cord sustain primitive characteristics during extensive expansion. Cell Tissue Res. 2008; 334(3): 423–433. PubMed Abstract \| Publisher Full Text\n\nAllsopp R, Shimoda J, Easa D, et al.: Long telomeres in the mature human placenta. Placenta. 2007; 28(4): 324–327. PubMed Abstract \| Publisher Full Text\n\nPacini S, Carnicelli V, Trombi L, et al.: Constitutive expression of pluripotency-associated genes in mesodermal progenitor cells (MPCs). PLoS One. 2010; 5(3): e9861. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu LL, Liu YJ, Yang SG, et al.: Isolation and characterization of human umbilical cord mesenchymal stem cells with hematopoiesis-supportive function and other potentials. Haematologica. 91(8): 1017–1026. PubMed Abstract\n\nBaird DM, Rowson J, Wynford-Thomas D, et al.: Extensive allelic variation and ultrashort telomeres in senescent human cells. Nat Genet. 2003; 33(2): 203–207. PubMed Abstract \| Publisher Full Text\n\nStenderup K, Justesen J, Clausen C, et al.: Aging is associated with decreased maximal life span and accelerated senescence of bone marrow stromal cells. Bone. 2003; 33(6): 919–26. PubMed Abstract \| Publisher Full Text\n\nBaxter MA, Wynn RF, Jowitt SN, et al.: Study of telomere length reveals rapid aging of human marrow stromal cells following in vitro expansion. Stem Cells. 2004; 22(5): 675–682. PubMed Abstract \| Publisher Full Text\n\nBernardo ME, Zaffaroni N, Novara F, et al.: Human bone marrow-derived mesenchymal stem cells do not undergo transformation after long-term In vitro culture and do not exhibit telomere maintenance mechanisms. Cancer Res. 2007; 67(9): 9142–9149. PubMed Abstract \| Publisher Full Text\n\nDe Bari C, Dell’Accio F, Karystinou A, et al.: A biomarker-based mathematical model to predict bone-forming potency of human synovial and periosteal mesenchymal stem cells. Arthritis Rheum. 2008; 58(1): 240–250. PubMed Abstract \| Publisher Full Text\n\nKim J, Kang JW, Park JH, et al.: Biological characterization of long-term cultured human mesenchymal stem cells. Arch Pharm Res. 2009; 32(1): 117–126. PubMed Abstract \| Publisher Full Text\n\nAviv A, Hunt SC, Lin J, et al.: Impartial comparative analysis of measurement of leukocyte telomere length/DNA content by Southern blots and qPCR. Nucleic Acids Res. 2011; 39(20): e134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNawrot TS, Staessen JA, Gardner JP, et al.: Telomere length and possible link to X chromosome. Lancet. 2004; 363(9408): 507–510. PubMed Abstract \| Publisher Full Text\n\nNordfjall K, Svenson U, Norrback KF, et al.: Large-scale parent-child comparison confirms a strong paternal influence on telomere length. Eur J Hum Genet. 2009; 18(3): 385–389. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFriedrich U, Griese E, Schwab M, et al.: Telomere length in different tissues of elderly patients. Mech Ageing Dev. 2000; 119(3): 89–99. PubMed Abstract \| Publisher Full Text\n\nTakubo K, Izumiyama-Shimomura N, Honma N, et al.: Telomere lengths are characteristic in each human individual. Exp Gerontol. 2002; 37(4): 523–531. PubMed Abstract \| Publisher Full Text\n\nGadalla SM, Cawthon R, Giri N, et al.: Telomere length in blood, buccal cells, and fibroblasts from patients with inherited bone marrow failure syndromes. Aging (Albany NY). 2010; 2(11): 867–74. PubMed Abstract \| Free Full Text\n\nKouroupis D, Churchman SM, McGonagle D, et al.: Data for CD146-based cell sorting and telomere length in umbilical cord. F1000Research. Data Source" }	[ { "id": "5083", "date": "04 Jul 2014", "name": "Armin Attar", "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 current study, the investigators have used cell sorting techniques as well as telomere length analysis to assess the identity of UC tissue derived mesenchymal stromal cells. The manuscript is scientifically sound and deserves indexation.", "responses": [] }, { "id": "5441", "date": "23 Jul 2014", "name": "Simone Pacini", "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 carefully read the manuscript entitled “The assessment of CD146-based cell sorting and telomere length analysis for establishing the identity of mesenchymal stem cells in human umbilical cord” from Kouropis et al. The aim of the study is verifying the hypothesis that umbilical cord-derived MSCs reside in the CD146-positive fraction, and could be discriminated by their possible longer telomeres. The authors demonstrated that frequency of CD146-positive cells in the culture-initiating suspension was much higher than the frequency of CFU-F confirming that this population also include non MSC-progenitors. Telomere length analysis revealed non-significant differences between CD146-positive and –negative cells, in three of four samples. Thus, the authors correctly conclude that the CD146-positive population was unlikely representing pure native UC-MSCs, and that a prospective isolation of ex vivo MSC necessarily should include more antigens to investigate. My review will be concentrated on some technical aspects regarding conduction of cell sorting and telomere analysis, which should be clarified. My major comments are as follows: From the density plots reported in figure 1A, it is not clear if a morphology gate was applied in the SSC/FSC plot, to exclude events of no-cellular origins. A Gate was applied in the dead-cells exclusion plot and in R1 would be counted no-cellular 7-AAD negative events, which could significantly alter the CD146 positive or negative percentages on total living cells. The authors extensively discussed the percentage of CD146-positive frequency in correlation to CFU-F frequency. Thus, the authors should clarify this aspect, and I strongly suggest clearing the flow cytometry data from the non-cellular events. As the CD146-positive and -negative populations are not sufficiently separated in the R4 gate, the sorting gates R5 and R6 should be defined by applying isotypic control. Moreover, I suggest reporting the isotypic control plot in the figure, in order to demonstrate the absence of events in the R5 region. Purity, and possible Yield, of sorted cells should be indicated in the results section, in order to reassure the reader of the quality of the sorting process. Telomere length has been described as T/S ratio, reported data are close to the value of 1.0 and two of four have values lower than 1.0. How could it be possible to detect telomere repeat copy numbers lower than the single copy gene (36B4)? According to Cawthon R.M. (2002) T/S ratios should be normalized. PCR efficiency could vary gene to gene, and should be evaluated in order to normalize qPCR data with the efficiency correction factor.", "responses": [ { "c_id": "948", "date": "21 Aug 2014", "name": "Dimitrios Kouroupis", "role": "Author Response", "response": "“From the density plots reported in figure 1A, it is not clear if a morphology gate was applied in the SSC/FSC plot, to exclude events of no-cellular origins. A Gate was applied in the dead-cells exclusion plot and in R1 would be counted no-cellular 7-AAD negative events, which could significantly alter the CD146 positive or negative percentages on total living cells. The authors extensively discussed the percentage of CD146-positive frequency in correlation to CFU-F frequency. Thus, the authors should clarify this aspect, and I strongly suggest clearing the flow cytometry data from the non-cellular events.” We are very grateful to Dr Pacini for his careful reading of our work and for his insightful comments. We believe that Figure 1A that he refers to is in fact Figure 2A as this pertains to our cell sorting experiments from umbilical cord (UC) tissue digests. We can confirm that we did perform FSC/SSC gating to remove anuclear red cells and cellular debris in all our sorting experiments. It was an oversight that this was omitted in the original submission and we thank the reviewer for highlighting this. We have now modified the relevant section of our Materials and Methods as well as Figure 2A to reflect this (gate R1). Additionally, we have added average percentage of cells within R1 out of all events collected into the appropriate Figure Legend. “As the CD146-positive and -negative populations are not sufficiently separated in the R4 gate, the sorting gates R5 and R6 should be defined by applying isotypic control. Moreover, I suggest reporting the isotypic control plot in the figure, in order to demonstrate the absence of events in the R5 region. Response: We appreciate this comment. In our experiments with UC tissue digests, a distinct CD45-CD235a-CD31-CD146+ population was always evident above the isotype control staining.” We further acknowledge that our original density plots failed to show that our sorting gates were always set apart to ensure maximal sorting purity. We have made new flow cytometry plots for panel 2A to reflect all of these changes (FSC/SSC gating and isotype control panel). “Purity, and possible Yield, of sorted cells should be indicated in the results section, in order to reassure the reader of the quality of the sorting process.”We have now provided information on CD146+ and CD146- fractions’ cell yields in Figure 2B, left panel. We acknowledge that we did not make clear in our original submission how many sorting experiments we performed in total, we confirm that this is ten for the sorting experiments. Fraction purities post-sort are always >95% using our cell sorter (see Churchman et al., 2012). “Telomere length has been described as T/S ratio, reported data are close to the value of 1.0 and two of four have values lower than 1.0. How could it be possible to detect telomere repeat copy numbers lower than the single copy gene (36B4)? According to Cawthon R.M. (2002) T/S ratios should be normalized. PCR efficiency could vary gene to gene, and should be evaluated in order to normalize qPCR data with the efficiency correction factor.”We appreciate this comment and agree that it is indeed a valid point that we should clarify further. Our method was based upon that of Cawthon R.M. (2002) as stated above and in the text, however when using unexpanded cells sorted from UC the cell numbers were limited because of the amount of tissue processed and inevitable cell losses caused by multiple centrifugation steps prior to cell sorting. Due to this we elected to standardise our telomere length experiments to 1000 sorted cells per fraction (previously overlooked from methods). We accept that we were working at the lower limit of DNA concentrations (acknowledgement of this is now added as a new sentence in Discussion), but believe that with triplicate replicates and extremely comparable standard curves (prepared using immortal cell lines) on each plate for inter-plate comparisons, our work is technically sound, and that our standardisation and technical control was suitable for comparing the differences between the CD146+ and CD146- populations. We did not normalise our T/S ratios and chose to show the data as a direct comparison of sample pairs and therefore some of these values fell below 1. This was also observed in the Cawthon paper; a ‘raw’ T/S ratio of 0.69 is what he normalised against to give the lowest value of 1 (Figure legend 5). Values below 1 can also be observed in Guillot et al. (2007).We have added an interesting observation that our umbilical cord T/S ratios fall closer to the adult tissues than the foetal tissues in that study. We now present normalized T/S ratios adjusted according to Cawthon R.M. (2002) in Figure 2C and conversion to telomere length in the text. We have also strengthened this figure with new data from a further 3 sorting experiments (new total n=7) and have performed statistical analysis. The CD146+ fraction possesses telomeres with median value 28bp longer than the CD146- fraction, although this was not significant (Wilcoxon)." } ] } ]	1	https://f1000research.com/articles/3-126
https://f1000research.com/articles/3-204/v1	27 Aug 14	{ "type": "Research Article", "title": "Difficult mask ventilation in general surgical population: observation of risk factors and predictors", "authors": [ "Davide Cattano", "Peter V. Killoran", "Chunyan Cai", "Anastasia D. Katsiampoura", "Ruggero M. Corso", "Carin A. Hagberg", "Peter V. Killoran", "Chunyan Cai", "Anastasia D. Katsiampoura", "Ruggero M. Corso", "Carin A. Hagberg" ], "abstract": "Background: There are few predictors of difficult mask ventilation and a simple, objective, predictive system to identify patients at risk of difficult mask ventilation does not currently exist. We present a retrospective - subgroup analysis aimed at identifying predictive factors for difficult mask ventilation (DMV) in patients undergoing pre-operative airway assessment before elective surgery at a major teaching hospital.Methods: Data for this retrospective analysis were derived from a database of airway assessments, management plans, and outcomes that were collected prospectively from August 2008 to May 2010 at a Level 1 academic trauma center. Patients were stratified into two groups based on the difficulty of mask ventilation and the cohorts were analyzed using univariate analysis and stepwise selection method.Results: A total of 1399 pre-operative assessments were completed with documentation stating that mask ventilation was attempted. Of those 1399, 124 (8.9%) patients were found to be difficult to mask ventilate. A comparison of patients with and without difficult mask ventilation identified seven risk factors for DMV: age, body mass index (BMI), neck circumference, history of difficult intubation, presence of facial hair, perceived short neck and obstructive sleep apnea. Although seven risk factors were identified, no individual subject had more than four risk factors.Conclusion: The results of this study confirm that in a real world clinical setting, the incidence of DMV is not negligible and suggest the use of a simple bedside predictive score to improve the accuracy of DMV prediction, thereby improving patient safety. Further prospective studies to validate this score would be useful.", "keywords": [ "airway management", "preoperative airway assessment", "difficult mask ventilation" ], "content": "Introduction\n\nOne of the primary responsibilities of an anesthesiologist is to maintain adequate oxygenation and ventilation by maintaining a patent upper airway1. Being able to provide ventilation by bag-mask, in face of a difficult or failed tracheal intubation, can make the difference between serious complications and disability.\n\nIn the literature, the incidence of difficult mask ventilation (DMV) varies from 0.08% to 15% depending on the definition used1–4. Despite its importance, there are few predictors of DMV2–4 and a simple, objective, predictive score to identify patients at risk of DMV at the bedside does not currently exist.\n\nWe present a retrospective-subgroup analysis of patients undergoing preoperative airway assessment before elective surgery5 at a major teaching hospital, to identify predictive factors for DMV and evaluate a composite score value, based on a comprehensive airway assessment and recorded outcomes.\n\n\nMaterials and methods\n\nData for this retrospective analysis were derived from a database of airway assessments, management plans, and outcomes collected prospectively from August 2008 to May 2010 at a Level 1 academic trauma center (Memorial Hermann Hospital, Texas Medical Center, Houston, TX, USA). The study was sponsored by an educational grant from the Foundation for Anesthesia, Education and Research (FAER), and other educational funds from the Department of Anesthesiology at University of Texas Medical School at Houston. After obtaining IRB approval (HSC-MS-07-0144) all non-obstetric adult patients presenting for elective surgery requiring general anesthesia, were enrolled in this study5.\n\nA total of 91 residents were involved in the data collection process. Residents were randomized into two groups — an experimental group of residents who used a comprehensive airway assessment form in addition to the existing anesthesia record, and a control group, who only used the existing anesthesia record. For the purpose of the present analysis, only the experiment group data were used for a total of 1339 recorded and attempted bag mask ventilations, graded and assigned to a pre-operative airway assessment5.\n\nDMV was defined as difficulty in maintaining a mask seal and obtaining a satisfactory capnography (end-tidal CO2 and tidal volume)6. If mask ventilation was attempted, then its easiness was determined and graded based on a severity score: from easy = 0, oral airway used = 1; to difficult, two handed ventilation = 2, or extraglottic device required = 3. However, the use of neuromuscular blocking agent, type, dosage, time of administration, and rescue was not considered in the analysis.\n\nDescriptive statistics mean ± standard deviation for continuous variables and frequency (percentage) for categorical variables was summarized for all pre-operative patient characteristics. Univariate analysis of comparison between patients with or without DMV was performed using the two sample t-test for continuous variables and the Chi-square test or Fisher exact test for categorical variables. In addition, receiver operating characteristic (ROC) curves were used to assess the discrimination ability of predicting DMV using continuous variables and to determine their best thresholds which maximize the sum of sensitivity and specificity. All dichotomized variables with a p-value <0.10 in univariate analysis were entered into a multivariate logistic regression model. A stepwise selection method was used to identify independent predictors of difficult mask ventilation. The adjusted odds ratios and their 95% confidence intervals (CI) were reported for each independent predictor. The area under a ROC curve or c-statistic was calculated to evaluate the resulting model’s predictive value.\n\nA non-weighted risk score was created by assigning one point to each independent predictor. In addition, a weighted score introduced in Kheterparl et al. (2009) was derived based on the coefficients of independent predictors from the logistic regression model8. The comparison between non-weighted and weighted risk scores was evaluated through c-statistic. All statistical analyses were conducted using SAS 9.3 (SAS Institute, Cary, NC, USA). A p-value <0.05 was considered significant.\n\n\nResults\n\nA total of 1399 pre-operative assessments were completed with documentation that MV was attempted, an ultimate outcome was graded, and the record was linked to a pre-operative airway assessment. Of 1399 patients, 124 (8.9%) were found to be difficult to mask ventilate (2 and 3, Table 1). Once stratified into two groups based on the difficulty of mask ventilation the cohorts were analyzed.\n\n* Mask ventilation was considered easy for MVEase classes 0 and 1 and difficult for MVEase classes 2 and 3. Local practice patterns often include placement of an oral airway for routine bag mask ventilation.\n\nBased on univariate analysis (Table 2), a total of eight factors were identified with a p-value <0.05: age, gender, BMI, neck circumference, history of difficult intubation, presence of facial hair, perceived short neck and obstructive sleep apnea (OSA, suspected or diagnosed). The thresholds that maximized the sum of sensitivity and specificity were 47 (year) for age, 35 (kg/m2) for BMI, and 40 (cm) for Neck Circumference by analyzing the ROC curve of each continuous risk factor to predict DMV. In addition to these significant factors, an additional variable capturing the absence of dentition (p=0.09) was included in the subsequent analysis. Entering all these nine factors into a multivariate logistic regression model, seven independent risks factors for DMV were identified using stepwise selection: age of 47 year or older, BMI of 35 kg/m2 or greater, and neck circumference of 40 cm or higher, history of difficult intubation, presence of facial hair, perceived short neck, and OSA; p<0.001; (Table 3). The model’s c-statistic is 0.75 (95% CI: 0.71-0.79), demonstrating a good discriminating capacity. The adjusted odds ratios are also presented in Table 3.\n\nNR: not reported due to zero cells. Values are reported as mean±SD and frequency (percentage).\n\nThe seven independent risk factors identified were then applied to all cases where DMV was encountered to evaluate a predictive model for DMV. Although seven risks factors were identified, no individual subject had more than four risk factors. As indicated, non-weighted and weighted risk score were created based on these seven risk factors. The model’s c-statistic based on unweighted score is 0.70 (95% CI: 0.66-0.74) (Figure 1). Weighted score did not improve the prediction performance, which model’s c-statistic is 0.70 (95% CI: 0.66-0.75). Therefore, we adopted the simple approach of unweighted risk score for the following analysis. The sensitivity, specificity, likelihood ratios, and predictive values were progressively calculated for patients with different number of risk factors (Table 4). The best cut-off for the number of risk factors was 2, which maximizes Youden’s index16 with sensitivity of 0.65 and specificity of 0.67. Table 5 also shows the distribution frequencies of different number of risk factors and the odds ratio for patients with one, two, or three risk factors relative to a patient with zero risk factors. When compared with zero risk factors, patients with two or more risk factors have an odds ratio of 7.6 (95% CI: 3.4-16.9).\n\nSeven independent predictors for difficult mask ventilation were identified using logistic regression: age of 47 yr or older, BMI of 35 kg/m2 or greater, NeckCirc of 40 or greater, HxDiffIntub, FacHair, short neck and OSA. A risk score for DMV was calculated based on the number of these seven risk factors a patient possessed. The area under the curve was 0.70±0.02.\n\nLikelihood ratio positive=Sensitivity/(1-Specificity)\n\nLikelihood ratio negative=(1-Sensitivity)/Specificity\n\nN/A: not applicable\n\n\nDiscussion\n\nFor more than three decades, poor airway management was recognized as a serious patient safety concern, emphasizing the need for a careful airway assessment and identifying the predictors for a difficult airway6. Moreover, the airway risk assessment tools in widespread use were mostly focused on one specific aspect of a difficult airway (i.e. difficult laryngoscopy, difficult intubation). In more recent years, this paradigm has shifted to a more functional approach with greater emphasis placed on the overall importance of the airway patency. Indeed, due to early data demonstrating the significant risk of respiratory depression associated with sedation, The Joint Commission and Centers for Medicare and Medicaid Services has implemented policies to ensure evaluation of the risk for a difficult airway prior to procedures. Moreover, the 2013 American Society of Anesthesiology (ASA) Practice Guidelines for Management of the Difficult Airway caution about the risks of a difficult bag-mask ventilation due to upper airway obstruction and recommend an airway risk assessment before every anesthesia procedure is performed7. In this study, we determine that: (1) the reported incidence of DMV was 9%; (2) the reported incidence of DMV in patients with a history of OSA was 17%; (3) seven independent risk factors were identified (age ≥ 47 yr, BMI ≥ 35 kg/m2, neck circumference ≥ 40 cm, history of difficult intubation, presence of facial hair, perceived short neck, history of OSA); (4) the absence of three of these factors allows to reasonably exclude a DMV (likelihood ratio negative: 0.85).\n\nRecent investigations have demonstrated that the incidence and risk factors for DMV are distinct from difficult laryngoscopy (DL) predictors (incidence i.e. ranges from 1.4%8 to 16%9). There are many reasons that can explain these findings: (1) absence of a universally accepted definition of DMV (different definitions lead to different data); (2) obesity and OSA are undoubtedly predictors of DMV, therefore a study done on a population with a high prevalence of obesity will show a higher incidence of DMV from a population with a lower prevalence of obesity; (3) the design of face masks and the technique used are not usually reported, but recent studies highlight their importance for performance and accordingly the reported incidence of DMV10,11; (4) the influence of neuromuscular block on mask ventilation has been demonstrated, but often these data are missing12.\n\nWe confirmed many factors such as age, short neck, facial hair, BMI, but most importantly neck circumference, that have been associated with difficult airway in the obese13,14 as well as a history of OSA. Interestingly, neck circumference and BMI are also important determinants for OSA screening, which may results in some overlap between OSA and DMV. Our study confirmed that OSA patients are at risk for DMV, calling for a systematic screening for OSA with the aim to identify a category of patients at risk of not only difficult airway, but also of post-operative complications15.\n\nWe attempted, indeed, to define a bedside score to predict DMV: our score has the advantage of including objective variables, such as neck circumference, but has a high false positive rate, possibly limiting the usefulness for a large-scale clinical implementation of the score. However with a sensitivity of 92% (using one risk factor, while it drops at lower values for two or more combined risk factors), this score could actually be useful as screening tool, since avoiding the underestimation of unpredicted DMV is far more important than a false positive (particularly in airway management where there are not significant costs attributed to overestimation).\n\nOur study also has other limitations: first, only DMV outcomes were analyzed without consideration for difficult laryngoscopy; second, a large number of records were selectively removed from our analysis because the outcomes were not known, reducing our statistical power and introducing the possibility of selection bias.\n\nThe results of this study confirm that in a real world clinical setting, the incidence of DMV is not negligible and suggest the use of a simple bedside predictive score to improve the accuracy of DMV prediction, thereby improving patient safety. Further prospective studies to validate this score would be useful.\n\n\nData availability\n\nData have been obtained from databases at the Memorial Hermann Hospital, Texas Medical Center, Houston, IRB approval HSC-MS-07-0144. The author can support applications to the Institutional Board to make the data accessible upon individual request.", "appendix": "Author contributions\n\n\n\nCattano: design, study monitoring, data interpretation, manuscript writing.\n\nChunyan: data quality, statistical analysis, data interpretation, manuscript writing.\n\nCorso: data interpretation, manuscript writing.\n\nKilloran: data quality, data interpretation.\n\nKatsiampoura: manuscript preparation.\n\nHagberg: design, manuscript preparation.\n\nAll authors read and approved the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe present work was supported by internal funds of the Department of Anesthesiology, University of Texas Medical School at Houston and the Foundation for Anesthesia Research, and Education (FAER). Dr. Killoran was supported by a training fellowship from the Keck Center NLM Training Program in Biomedical Informatics of the Gulft Coast Consortia (NLM grant no. T15LM007093). Dr. Cai’s research was supported by the National Institutes of Health’s Clinical and Translational Science Award grant (UL1 TR000371), awarded to the University of Texas Health Science Center at Houston in 2012 by the National Center for Clinical and Translational Sciences. Dr. Cattano receives grant support from Karl Storz Endoskope Inc. Dr. Hagberg was the recipient of the 2007 Foundation for Anesthesia Education and Research (FAER) Grant and currently receives grant support from AMBU, Covidien, and Karl Storz Endoskopy. She serves on the speaker bureau for Covidien, LMA North America, and Ambus A/S.\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\nEl-Orbany M, Woehlck HJ: Difficult mask ventilation. Anesth Analg. 2009; 109(6): 1870–80. PubMed Abstract \| Publisher Full Text\n\nLangeron O, Masso E, Huraux C, et al.: Prediction of difficult mask ventilation. Anesthesiology. 2000; 92(5): 1229–36. PubMed Abstract \| Publisher Full Text\n\nKheterpal S, Han R, Tremper K, et al.: Incidence and predictors of difficult and impossible mask ventilation. Anesthesiology. 2006; 105(5): 885–91. PubMed Abstract \| Publisher Full Text\n\nKheterpal S, Healy D, Aziz MF, et al.: Multicenter Perioperative Outcomes Group (MPOG) Perioperative Clinical Research Committee. Incidence, predictors, and outcome of difficult mask ventilation combined with difficult laryngoscopy: a report from the multicenter perioperative outcomes group. Anesthesiology. 2013; 119(6): 1360–9. PubMed Abstract \| Publisher Full Text\n\nCattano D, Killoran PV, Iannucci D, et al.: Anticipation of the difficult airway: preoperative airway assessment, an educational and quality improvement tool. Br J Anaesth. 2013; 111(2): 276–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeterson GN, Domino KB, Caplan RA, et al.: Management of the difficult airway: a closed claims analysis. Anesthesiology. 2005; 103(1): 33–9. PubMed Abstract \| Publisher Full Text\n\nApfelbaum JL, Hagberg CA, Caplan RA, et al.: American Society of Anesthesiologists Task Force on Management of the Difficult Airway. Practice guidelines for management of the difficult airway: an updated report by the American Society of Anesthesiologists Task Force on Management of the Difficult Airway. Anesthesiology. 2013; 118(2): 251–70. PubMed Abstract \| Publisher Full Text\n\nKheterpal S, Martin L, Shanks AM, et al.: Prediction and outcomes of impossible mask ventilation: a review of 50,000 anesthetics. Anesthesiology. 2009; 110(4): 891–7. PubMed Abstract \| Publisher Full Text\n\nRoberts S, Cyna AM, Walsh JP, et al.: Assessment of anaesthetists’ ability to predict difficulty of bag-mask ventilation. Br J Anaesth. 2013; 111(4): 676–7. PubMed Abstract \| Publisher Full Text\n\nNa JU, Han SK, Choi PC, et al.: Influence of face mask design on bag-valve-mask ventilation performance: a randomized simulation study. Acta Anaesthesiol Scand. 2013; 57(9): 1186–92. PubMed Abstract \| Publisher Full Text\n\nJoffe AM, Hetzel S, Liew EC: A two-handed jaw-thrust technique is superior to the one-handed “EC-clamp” technique for mask ventilation in the apneic unconscious person. Anesthesiology. 2010; 113(4): 873–9. PubMed Abstract \| Publisher Full Text\n\nWarters RD, Szabo TA, Spinale FG, et al.: The effect of neuromuscular blockade on mask ventilation. Anaesthesia. 2011; 66(3): 163–7. PubMed Abstract \| Publisher Full Text\n\nGonzalez H, Minville V, Delanoue K, et al.: The importance of increased neck circumference to intubation difficulties in obese patients. Anesth Analg. 2008; 106(4): 1132–6. PubMed Abstract \| Publisher Full Text\n\nLeoni A, Arlati S, Ghisi D, et al.: Difficult mask ventilation in obese patients: analysis of predictive factors. Minerva Anestesiol. 2014; 80(2): 149–57. PubMed Abstract\n\nCorso RM, Petrini F, Buccioli M, et al.: Clinical utility of preoperative screening with STOP-Bang questionnaire in elective surgery. Minerva Anestesiol. 2014; 80(8): 877–84. PubMed Abstract\n\nYouden WJ: Index for rating diagnostic tests. Cancer. 1950; 3(1): 32–35. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5963", "date": "28 Aug 2014", "name": "D John Doyle", "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 important contribution to the difficult airway literature and will be helpful to clinicians wanting a means to predict that mask ventilation may be a problem following the induction of general anesthesia. While the limitations of the study are correctly stated (see p. 5) I would add some wording to this section to re-emphasize that this was a retrospective study and that the use of neuromuscular blocking drugs was not part of the analysis. In addition, since a great many cases of difficult mask ventilation vanish following the administration of neuromuscular blocking drugs, I recommend that the discussion on this particular issue should be discussed in a bit more detail, perhaps making reference to one or more of the following recent publications.Xue FS, Cheng Y, Li RP: Facemask ventilation and neuromuscular blockade in anesthetized patients. Anesthesiology. 2013;118(4):991-2. doi: 10.1097/ALN.0b013e3182874628Richardson MG, Litman RS: Ventilation before paralysis: crossing the Rubicon, slowly. Anesthesiology. 2012; 117(3):456-8. doi: 10.1097/ALN.0b013e318266868fEngelhardt T, Weiss M: Difficult mask ventilation and muscle paralysis. Anesthesiology. 2013; 118(4):994. doi: 10.1097/ALN.0b013e3182874659.Priebe HJ: Ventilation before paralysis. Anesthesiology. 2013; 118(4):992-3. doi: 10.1097/ALN.0b013e3182874642.", "responses": [ { "c_id": "955", "date": "28 Aug 2014", "name": "Davide Cattano", "role": "Author Response", "response": "I am grateful to Dr Doyle for his review, comments and, in particular, for the detailed references.Indeed the avoidance of muscle relaxation from the analysis was a major \"missed\" undertaking, since you are completely right about the important effects of muscle relaxation. We made sure that such information was immediately available to the reader as a methodological point, and we kept it short in the discussion to focus on positive findings. It is always of importance to report both positive and negative findings; however, the lack of such related data was missing information rather than negative. We appreciate, however, your comment giving an opportunity for the reader to understand the importance of such a factor. We also need to remind the reader of the interesting aspect that muscle relaxation plays in difficult airway management and how it also plays a role in the DAW algorithms. It is not trivial to consider that our practice includes the establishment of an airway after a mask ventilation is achieved without the utilization of muscle relaxant at first (which we included as standard of care in our investigation) while muscle relaxant is utilized as a rescue for such events.It is important to consider the evidence value of certain practices though - whether or not they not only make sense, but are of value for a safe and practical airway management. In fact, the controversy - as you elegantly point at with the references provided - is about giving muscle relaxant up front (which was part of anesthesia practice a few years ago). Based on experience, the question is rather: can I use muscle relaxant or not, and should I do an awake intubation (or other airway)? Also the major change that occurred from the recognition of airway related morbidity and mortality in the 80's and 90's and now, is the availability of extraglottic devices and new airway tools that have in fact improved our practice." } ] }, { "id": "6308", "date": "17 Oct 2014", "name": "Adrian Matioc", "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 is today an increasing interest in studying difficult face mask ventilation. Cattano et al. present a retrospective study derived from a database built of elective cases. There are several points I want to make:First, the study used a “real world clinical setting” confirming that in the daily practice difficult face mask ventilation is not a rare occurrence.Second, it introduced an objective assessment tool for the difficult face mask ventilation: capnography (further comment from the authors regarding this tool would be appreciated).Third: six out of the seven independent risk factors for the difficult face mask ventilation (age, BMI, neck circumference, presence of facial hair, perceived short neck, OSA) that were identified and then applied to evaluate a predictive model can be easily assessed or inferred with a clinical exam on an unconscious patient thus having probably a role in the outside of the operating room setting too.Fourth: the study confirms the importance of neck circumference, BMI and OSA in the difficult face mask ventilation paradigm. It is known that these clinical settings are plagued by inspiratory and expiratory obstruction at the soft palate level that does not respond to airway manoeuvres. The soft palate obstruction (relevant for nasal ventilation) can be by passed by committing to oral ventilation (by opening the mouth with an oropharyngeal airway). 1 2 It may be appropriate to recommend in these clinical settings a first (optimized) face mask ventilation attempt with an oropharyngeal airway in situ.", "responses": [ { "c_id": "1040", "date": "17 Oct 2014", "name": "Davide Cattano", "role": "Author Response", "response": "We are grateful for Dr Matioc comments which endorse our findings while offring an opportunity for further discussion. We do agree with Dr Matioc about the importance of properly monitoring ventilation, which may not always be possible simply by utilizing a capnometry tool, and current investigations are actually looking at other type of devices, particularly in the postoperative arena, where airway obstruction may still occur. In the operating room the presence of capnometry combined with spirometry is a complete tool. Regarding airway adjuncts, I agree that an oral airway cannula should be promptly used, as in fact was one of the limitations of our investigation, since it is part of common airway management pathways, however I would consider also a nasal airway, if available and not concerning trauma, for a better and proper management of ventilation." } ] } ]	1	https://f1000research.com/articles/3-204
https://f1000research.com/articles/3-203/v1	27 Aug 14	{ "type": "Research Article", "title": "One year adjustable intragastric balloon: safety and efficacy of the Spatz3 adjustable balloons", "authors": [ "Evzen Machytka", "Jeffrey Brooks", "Marek Buzga", "John Mason", "Jeffrey Brooks", "Marek Buzga", "John Mason" ], "abstract": "Background: The Spatz3 Adjustable balloon system is approved for 1 year implantation and allows multiple changes in the balloon volume during the course of implantation.Other intragastric balloons are currently approved for 6 months and their balloon volumes cannot be adjusted after implantation.Aim: To determine the efficacy and safety of the Spatz3 adjustable balloon system.Methods: Seventy seven consecutive patients (66 females and 11 males) in two medical centers were implanted with the Spatz3 adjustable balloon device and were followed prospectively. The patients’ mean BMI was 37.2; the mean weight was 108.7 kg; the mean age was 41 (16-68); the mean balloon volume was 469 ml (450-500 ml). Adjustments were made for intolerance or weight loss plateau.Results: The mean weight loss at 1 year was 17.2 kg with 15.9 % weight loss and 42.9 % Excess Weight Loss (%EWL). Eighteen patients underwent balloon volume adjustments: three downward adjustments of 100 -150 cc which alleviated early intolerance; 15 upward adjustments (mean 320 ml; range 200-500) at a mean 4.1 months (range 2-5 months) yielded additional mean wt loss of 8.2 kg (range 0-25 kg) after the adjustment. Three balloons were removed before the 1 year completion date due to intolerance and three others were removed for other reasons (pregnancy, gall bladder surgery, and alcoholism). There was one episode of gastric ulceration which required endoscopic therapy and balloon removal. There were no deflations or perforations.\n\nConclusions: The Spatz3 adjustable balloon is a safe and effective treatment for weight loss. The adjustability function can yield greater weight loss for those who show weight loss plateau and can mitigate intolerance.", "keywords": [ "Obesity", "weight loss", "intragastric balloons", "adjustable gastric balloon", "weight loss plateau", "intolerance" ], "content": "Introduction\n\nIntragastric balloons (IGBs) have been successfully used to achieve weight loss for the past 30 years. Previously published results revealed an average weight loss of 12–15 kg over 6 months with an excellent safety profile1–16. The Spatz Adjustable balloon was introduced in 2010 as the first IGB approved for 1 year implantation while featuring an adjustability function that afforded balloon volume changes as needed. It was approved in May 2010 in all 27 countries of the European Union for patients with BMI > 27 that failed previous attempts at weight loss (certificate number 10 0384 QS/NB). The adjustability function was developed to address the issues related to other standard 6 months IGBs; 1) reduced efficacy after 2 to 3 months from implantation3,16, and 2) significant nausea, vomiting, and discomfort in the early implantation period, necessitating balloon extraction in 4–7% of patients1,2. Two studies have reported results showing that the Spatz adjustable balloon can result in weight losses of 21.6 kg (45.7% Excess Weight Loss, EWL) and 24.4 kg (48.8% EWL) respectively17,18. In 2012, the Spatz3 adjustable balloon was approved for 1 year implantation with the same features as the original Spatz balloon, but with a softer and smaller profile catheter.\n\nWe report our experiences with the Spatz3 adjustable balloon in the Czech Republic and the UK.\n\n\nPatients and methods\n\nThe Spatz3 ABS (Spatz3 Adjustable Balloon System, Spatz FGIA, Inc. NY, USA) was implanted at the University Hospital, Ostrava, Czech Republic and at the Trafford General Hospital, Manchester, UK. Seventy seven consecutive patients were selected according to National Institutes of Health criteria and guidelines for obesity surgery19 and were independently evaluated by members of staff: gastroenterologists, dieticians, and psychologists. Previous esophageal or gastric surgery, bowel strictures or history of bowel obstruction, inability to comply with frequent follow up, inability to tolerate endoscopic procedures or multiple episodes of vomiting were grounds for exclusion. No patients were excluded from implantation. The implantations were performed under the approval of the Ethics Committee and informed consent was obtained from all patients. The records of relevant co-morbidities, medications or family history were not available for review. Indications for Spatz ABS implantation included one of the following: (1) temporary weight loss treatment in patients with body mass index (BMI) in the range of bariatric surgery (>35) who refused surgery or were at high risk for surgery, (2) temporary weight loss treatment for patients with no indications for surgery (BMI 29–35). All patients underwent endoscopy using conscious sedation (Midazolam 5–10 mg and Fentanyl 50–100 mcg).\n\nBalloons were inflated with normal saline with the addition of 5 ml of undiluted methylene blue. Patients were recovered for 45 minutes and discharged the same day on pantoprazole 40 mg BID, ondansetron 4 mg BID, and a progressive clear liquid diet for 3–5 days. After the fifth postoperative day, the patients began a progressive solid 1,000 kcal diet. Monthly follow up was offered to all patients after implantation. Patients who were intolerant to the balloon could be adjusted downward by 100–150 ml, and those who did not lose weight, or whose weight loss reached a plateau could be adjusted upward by a volume of 250–500 ml at the discretion of the endoscopist. After 12 months of placement, the balloon was deflated by grasping the valve with a snare, and attaching an extension tube through which suction was applied, or via needle puncture. The device was removed endoscopically under conscious sedation using a grasping forceps or a polypectomy snare.\n\n\nEndpoints\n\nThe primary endpoints of the study were successful implantation, adjustment and extraction of the Spatz device, without bowel obstruction, perforation, ulceration or hemorrhage. Secondary endpoints were a > 10% weight loss; additional weight loss following adjustment; and salvage patients following adjustment for intolerance. The final overall results were calculated based on all patients that were implanted with the balloon, including those that completed treatments as well as those that did not complete 12 months for various reasons. Those patient dropout results were calculated based on weight at 12 months.\n\n\nResults\n\nFrom January 2012 to June 2013, 77 consecutive patients (66 female, 11 male) underwent Spatz3 ABS placement with demographics displayed in Table 1. The mean age was 41, with a mean weight of 108.7 kg, and a mean excess weight of 42.9 kg and a mean BMI of 37.2.\n\nSeventy patients completed the 12 month implantation period and seven patients underwent early removal. The causes of early removal were pregnancy, gall bladder surgery, alcoholism, gastroesophageal reflux disease (GERD), early intolerance and intermittent intolerance (both refusing downward adjustment) and one bleeding gastric ulcer. These are displayed in Table 2.\n\nThe mean weight loss was 17.2 kg, with a mean % weight loss of 15.9% and a mean % excess weight loss of 40.1%. A > 10% weight loss was obtained in 80.3% of our patients.\n\nWeight loss results of the implanted patients are displayed in Table 3.\n\n\nAdjustments\n\nSix patients were intolerant to the balloon beyond 1 week. Three of them agreed to volume adjustment and were able to complete the 12 month treatment period. The other three patients refused downward adjustment and the balloons were extracted prematurely at 3 ½ weeks, 2 months and 6 ½ months, respectively (Table 2). The mean of additional weight loss after downward adjustment was 12.7 kg (Table 4).\n\nFifteen patients underwent upward adjustments as weight loss reached a plateau. All adjustments were successful and resulted in a further mean weight loss of 8.2 kg. The extra weight loss that resulted after adjustment ranged from zero to 25 kg (0, 1, 3.2, 4, 4, 5, 5.9, 6, 7, 8, 11, 12, 13, 18, 25 kg).\n\n\nComplications\n\nComplications included intolerance in six patients, of which three agreed to downward adjustment resulting in alleviation of intolerance – the other three patients had their balloons removed and remained asymptomatic post extraction. During one adjustment procedure the valve disconnected requiring a balloon replacement – the patient continued with the new balloon to the end of the 12 month implantation period. There was one gastric ulceration at 5 months which required endoscopic therapy for bleeding and balloon removal. The patient was discharged post extraction of the balloon on pantoprazole 40 mg BID and remained asymptomatic with confirmed healing of the ulcer on subsequent endoscopy. There were no deflations or perforations.\n\n\nDiscussion\n\nHaving previous experience with the original Spatz Adjustable balloon from 2010 to 2011 in over 180 procedures in our two institutions, it is our opinion that the original Spatz had a lengthier and more complicated implantation and extraction procedures, whereas the procedures for Spatz3 are less complicated with fewer steps. The weight loss results are comparable with respect to % EWL (45% and 48% in Spatz and 40.7% EWL in Spatz3). The complication rate reported with the original Spatz was 4.1% with a deflation rate of 4% which has diminished with Spatz3- (1.3% complication rate and no deflations)17,18. We can conclude that the Spatz3 balloon is easier to use and has the same efficacy with a very low side effect profile. We report our experience with the Spatz3 ABS in 77 consecutive patients. In the present experience, other than intolerance in six patients and one bleeding gastric ulcer requiring endotherapy and balloon removal, there were no other complications. Three of the 6 intolerant patients agreed to downward adjustments which alleviated their symptoms. The additional mean weight loss of 12.7 kg is similar to the previously reported 13.2 kg lost after downward adjustment18.\n\nOur experience with upward adjustments confirms earlier reports of added weight loss after adjustment. We report a mean additional weight loss after upward adjustment of 8.2 kg which is similar to the 9.4 kg reported in the previous Spatz publication18. Adjustment was an effective tool for our intolerant patients as well as for our patients with weight loss plateau who were willing to undergo an extra procedure. In the Czech Republic, all of our patients except one underwent upward adjustment whereas in the UK only 12% of our patients opted for upward adjustment. We believe this is as a result of the long distance travel for our UK patients from all over the UK to the UK center in Manchester.\n\nBased on our experience, the Spatz3 Adjustable Balloon System is an effective procedure for weight reduction, without mortality, and very limited morbidity.\n\n\nData availability\n\nF1000Research: Dataset 1. Data on safety and efficacy of Spatz3 adjustable balloons, 10.5256/f1000research.5099.d3485020", "appendix": "Author contributions\n\n\n\nDr Machytka and Dr Mason submitted the data that was analyzed and presented in the paper. Dr Brooks reviewed the final manuscript and made suggestions prior to submission by Dr Machytka. All authors agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nDr Jeffrey Brooks is the CEO of Spatz FGIA Inc which is the manufacturer of the Spatz balloon. Dr Evzen Machytka and Dr John Mason have no conflict of interest to report.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMathus-Vliegen EM: Intragastric balloon treatment for obesity: what does it really offer? Dig Dis. 2008; 26(1): 40–4. PubMed Abstract \| Publisher Full Text\n\nImaz I, Martínez-Cervell C, García-Alvarez EE, et al.: Safety and effectiveness of the intragastric balloon for obesity. A meta-analysis. Obes Surg. 2008; 18(7): 841–6. PubMed Abstract \| Publisher Full Text\n\nBonazzi P, Petrelli MD, Lorenzini I, et al.: Gastric emptying and intragastric balloon in obese patients. Eur Rev Med Pharmacol Sci. 2005; 9(5 Suppl 1): 15–21. PubMed Abstract\n\nMion F, Napoléon B, Roman S, et al.: Effects of intragastric balloon on gastric emptying and plasma ghrelin levels in non-morbid obese patients. Obes Surg. 2005; 15(4): 510–6. PubMed Abstract \| Publisher Full Text\n\nEvans JD, Scott MH: Intragastric balloon in the treatment of patients with morbid obesity. Br J Surg. 2001; 88(9): 1245–8. PubMed Abstract \| Publisher Full Text\n\nSallet JA, Marchesini JB, Paiva DS, et al.: Brazilian multicenter study of the intragastric balloon. Obes Surg. 2004; 14(7): 991–8. PubMed Abstract \| Publisher Full Text\n\nGenco A, Cipriano M, Bacci V, et al.: BioEnterics Intragastric Balloon (BIB): a short-term, double-blind, randomized, controlled, crossover study on weight reduction in morbidly obese patients. Int J Obes (Lond). 2006; 30(1): 129–33. PubMed Abstract \| Publisher Full Text\n\nDoldi SB, Micheletto G, Perrini MN, et al.: Intragastric balloon: another option for treatment of obesity and morbid obesity. Hepatogastroenterology. 2004; 51(55): 294–7. PubMed Abstract\n\nRoman S, Napoleon B, Mion F, et al.: Intragastric balloon for “non-morbid” obesity: a retrospective evaluation of tolerance and efficacy. Obes Surg. 2004; 14(4): 539–44. PubMed Abstract \| Publisher Full Text\n\nDoldi SB, Micheletto G, Di Prisco F, et al.: Intragastric balloon in obese patients. Obes Surg. 2000; 10(6): 578–81. PubMed Abstract \| Publisher Full Text\n\nAl-Momen A, El-Mogy I: Intragastric balloon for obesity: a retrospective evaluation of tolerance and efficacy. Obes Surg. 2005; 15(1): 101–5. PubMed Abstract \| Publisher Full Text\n\nHerve J, Wahlen CH, Schaeken A, et al.: What becomes of patients one year after the intragastric balloon has been removed? Obes Surg. 2005; 15(6): 864–70. PubMed Abstract \| Publisher Full Text\n\nMelissas J, Mouzas J, Filis D, et al.: The intragastric balloon—smoothing the path to bariatric surgery. Obes Surg. 2006; 16(7): 897–902. PubMed Abstract \| Publisher Full Text\n\nBusetto L, Segato G, De Luca M, et al.: Preoperative weight loss by intragastric balloon in super-obese patients treated with laparoscopic gastric banding: a case-control study. Obes Surg. 2004; 14(5): 671–6. PubMed Abstract \| Publisher Full Text\n\nDoldi SB, Micheletto G, Perrini MN, et al.: Treatment of morbid obesity with intragastric balloon in association with diet. Obes Surg. 2002; 12(4): 583–7. PubMed Abstract \| Publisher Full Text\n\nTotte E, Hendrickx L, Pauwels M, et al.: Weight reduction by means of an intragastric device; experience with Bioenterics Intragastric Balloon. Obes Surg. 2001; 11(4): 519–23. PubMed Abstract \| Publisher Full Text\n\nMachytka E, Klvana P, Kornbluth A, et al.: Adjustable intragastric balloons: a 12-month pilot trial in endoscopic weight loss management. Obes Surg. 2011; 21(10): 1499–507. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrooks J, Srivastava ED, Mathus-Vliegen EM: One-year adjustable intragastric balloons: results in 73 consecutive patients in the U.K. Obes Surg. 2014; 24(5): 813–819. PubMed Abstract \| Publisher Full Text\n\nNIH Conference. Gastrointestinal surgery for severe obesity. Consensus development conference panel. Ann Intern Med. 1991; 115(12): 956–61. PubMed Abstract \| Publisher Full Text\n\nMachytka E, Brooks J, Buzga M, et al.: Data on safety and efficacy of Spatz3 adjustable balloons. F1000Research. 2014. Data Source" }	[ { "id": "6789", "date": "24 Nov 2014", "name": "Marko Nikolic", "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 have analyzed the efficacy of adjustable intragastric balloon in mid-term weight loss. There are few questions which need to be raised.How many patients have achieved success after 6 and 12 months of treatment (defined as EXWL > 20%). It would be of reader’s interest to know EXWL in first 6 months and in the next six months. Then we could easily compare the results with BIB. How many patients experienced nausea, dyspeptic symptoms etc? In the discussion section, please compare the efficacy and side effects between Spatz1 and Spatz3. Additionally, compare Spatz3 with BIB (since BIB is the gold standard of treatment with intragastric balloons). Please mention how long did it take to perform additional adjustments, how many additional visits were necessary and what were the complications of all adjustments. All the above required changes should be presented in new tables or figures, since the current ones are too small and hard to follow.", "responses": [] }, { "id": "7945", "date": "20 Mar 2015", "name": "Jamie Ponce", "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\nOverall good series with an adjustable intragastric balloon.In the Methods, need to describe the initial fluid volume Describe the parameters used to define “intolerance” as well as “weight plateau” for removing or adding fluid. Need to present weight loss at 6 month interval as well to compare with 6-month balloons data, also will tell us how much effect the balloon has on the 2nd 6-month period. May need to explain if there were similar features on pts requiring upward adjustments (i.e., height, stomach size, etc.) Discussion needs to include comparative analysis with other balloons data, or mention systematic review", "responses": [] } ]	1	https://f1000research.com/articles/3-203
https://f1000research.com/articles/3-151/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "Cytoscape: the network visualization tool for GenomeSpace workflows", "authors": [ "Barry Demchak", "Tim Hull", "Michael Reich", "Ted Liefeld", "Michael Smoot", "Trey Ideker", "Jill P. Mesirov", "Tim Hull", "Michael Reich", "Ted Liefeld", "Michael Smoot", "Trey Ideker", "Jill P. Mesirov" ], "abstract": "Modern genomic analysis often requires workflows incorporating multiple best-ofbreed tools. GenomeSpace is a web-based visual workbench that combines a selection of these tools with mechanisms that create data flows between them. One such tool is Cytoscape 3, a popular application that enables analysis and visualization of graph-oriented genomic networks. As Cytoscape runs on the desktop, and not in a web browser, integrating it into GenomeSpace required special care in creating a seamless user experience and enabling appropriate data flows. In this paper, we present the design and operation of the Cytoscape GenomeSpace app, which accomplishes this integration, thereby providing critical analysis and visualization functionality for GenomeSpace users. It has been downloaded it over 850 times since the release of its first version in September, 2013.", "keywords": [ "GenomeSpace is a web-based application (http://genomespace.org) that provides a workspace environment for executing biologic analysis workflows involving genomic data. It hosts a variety of third party tools that it can launch to perform queries of public genomic databases", "customized analyses", "and customized visualization and publishing. The data and results from a given tool are stored by GenomeSpace as private", "public", "or shared files that are made available to other tools in the workspace. Cytoscape is a standalone desktop application that enables users to analyze", "visualize", "and publish complex networks – with its GenomeSpace app (“the app”", "http://apps.cytoscape.org/apps/genomespace)", "it doubles as a tool available to GenomeSpace workflows. This paper describes Cytoscape’s GenomeSpace app", "which links GenomeSpace and Cytoscape", "thereby enabling data to flow between Cytoscape and other GenomeSpace tools." ], "content": "Introduction\n\nGenomeSpace is a web-based application (http://genomespace.org) that provides a workspace environment for executing biologic analysis workflows involving genomic data. It hosts a variety of third party tools that it can launch to perform queries of public genomic databases, customized analyses, and customized visualization and publishing. The data and results from a given tool are stored by GenomeSpace as private, public, or shared files that are made available to other tools in the workspace. Cytoscape is a standalone desktop application that enables users to analyze, visualize, and publish complex networks – with its GenomeSpace app (“the app”, http://apps.cytoscape.org/apps/genomespace), it doubles as a tool available to GenomeSpace workflows. This paper describes Cytoscape’s GenomeSpace app, which links GenomeSpace and Cytoscape, thereby enabling data to flow between Cytoscape and other GenomeSpace tools.\n\nWhile there exists a number of workflow engines that can perform biological analyses (e.g., Taverna1, BioKepler2, and Galaxy3), GenomeSpace distinguishes itself by combining a collaboration-oriented file system (incorporating sharing and stored metadata), a robust and user-extensible spectrum of genomic tools, and a library of recipes demonstrating best practices for the orchestration of GenomeSpace tools to achieve common and important bioinformatic results. While some tools implement specific and constrained functionality (e.g., ISAcreator4), others are complex and rich applications (e.g., GenePattern5, Gitools6, and Cistrome7), and yet others are fully featured workflow management systems themselves (e.g., Galaxy and Sage Synapse8). By tying these features together, GenomeSpace provides a comprehensive and effective environment for genomic research.\n\nGenomeSpace offers a short list of tools that enable network visualization and analysis, including Cytoscape, Genomica (http://genomica.weizmann.ac.il/) (for module network trees), GiTools6 (for heat maps), and IGV9 (for sequencing data). Cytoscape distinguishes itself by being graph-oriented and delivering rich filtering, layout, and visual style features backed up by an extensive collection of third-party apps10, including pathway analysis, data integration, GO annotation, and more.\n\nTo launch a tool, GenomeSpace opens a new browser page using a URL specific to the tool. Paradoxically, Javabased tools (such as Cytoscape) run directly on the user’s workstation instead of within a browser. In this paper, we describe a specialized launch strategy that addresses this. We also describe Cytoscape’s GenomeSpace app, how its user interface adds GenomeSpace functionality within Cytoscape, and how it uses GenomeSpace’s Client Development Kit (CDK) to access the GenomeSpace file system to read input files or write result files.\n\nNote that GenomeSpace supports two Cytoscape tools it calls “Cytoscape” and “Cytoscape 3”. Its “Cytoscape” refers to the deprecated Cytoscape version 2, and its “Cytoscape 3” refers to Cytoscape version 3, which is the currently released version (http://cytoscape.org). Within this paper, we discuss only the “Cytoscape 3” tool, and refer to it simply as “Cytoscape”.\n\n\nImplementation\n\nThe Cytoscape support for GenomeSpace exists in three parts: the launch support, Cytoscape’s GenomeSpace app, and Cytoscape itself. This section describes how the launch support and app work, and leaves the operation of Cytoscape to the Results section below.\n\nTo launch a tool, a GenomeSpace user left-clicks on the corresponding toolbar icon, which activates the tool via the tool’s URL. For Cytoscape, the URL references a launch descriptor file that adheres to the Java Network Launch Protocol (JNLP)11 and resides on the Cytoscape web site. Web browsers are defined to process a JNLP file URL by starting a specialized launcher that downloads a Java application named in the JNLP file, then executes it on the user’s workstation. In Cytoscape’s case, we created a dynamic JNLP file (as a PHP script that delivers a JNLP file) that executes a small LaunchHelper Java application (see Figure 1). GenomeSpace constructs the JNLP URL to contain a GenomeSpace file descriptor as a parameter, and the PHP script extracts it and defines it as a parameter to the LaunchHelper (see Supplementary Data).\n\nThe LaunchHelper tests for the presence of Cytoscape and Cytoscape’s GenomeSpace app, and installs them if they are not present. Because LaunchHelper is itself a Java application, it is able to download the Cytoscape installer appropriate for the user’s workstation and the GenomeSpace app while maintaining an interactive user interface, including appropriate installation dialog boxes (as JOptionPane) and progress bars (as ProgressMonitorInputStream).\n\nNote that before attempting to launch Cytoscape, GenomeSpace attempts to determine if Cytoscape is already running by using the JRAC protocol (http://code.google.com/p/jrac) – if it is, Cytoscape simply starts a new session.\n\nThe GenomeSpace app manages the relationship between Cytoscape and GenomeSpace once Cytoscape is running. In addition to responding to a JRAC request (above), it augments the Cytoscape user interface to allow the user to directly access the GenomeSpace file system and tools.\n\nThe app exposes a number of GenomeSpace functions as menu items under Cytoscape’s File and Apps menus. This enables GenomeSpace login, opening and saving GenomeSpace sessions, launching GenomeSpace tools, and importing and exporting networks and tables as GenomeSpace files. Networks can be exported in .sif, .cyjs, .nnf, PSI-MI, and .xgmml formats. While some menu items are positioned within Cytoscape’s top-level menus (e.g., session open and save menu items in the File menu), others are positioned in submenus within Cytoscape menu items (e.g., importing a network under File \| Import \| Network). For nice effect, each menu item identifies itself with a distinctive GenomeSpace logo and uses menu gravity to place itself consistently relative to existing Cytoscape menu items. (It uses setPreferredMenu to add the menu, setMenuGravity to position it, and putValue to set the small icon).\n\nTo implement these menu items, the app communicates with GenomeSpace via the GenomeSpace CDK (http://www.genomespace.org/support/api/cdk), a proxy interface to GenomeSpace carried over an SSL Internet connection. The CDK enables GenomeSpace session management, tool discovery, user authentication, file system listing, and file upload and download. As with other apps, the app’s cyActivator initializes the app state, including gaining references to the standard Cy objects: application, network, view, and table managers. It also initializes basic CDK-related state (e.g., the GenomeSpaceContext root context).\n\nIn addition to communicating with GenomeSpace, the CDK displays key GenomeSpace-related dialog boxes, including the login dialog and a file chooser for import and export functions. The app uses file choosers to identify an import (or export) file (referenced by metadata), but then executes the operation using a combination of CDK download/upload functions and Cytoscape task manager and monitor functions.\n\nFor example, given metadata for a network file to import, the app creates a Cytoscape task iterator that downloads the network to a temp file, loads the network into Cytoscape, and then deletes the temp file as shown in the Supplementary Data. DownloadFileFromGenomeSpaceTask calls CDK to perform the download, and loadNetworkFileTaskFactory adds the network to the Cytoscape data model. Because the app orchestrates the download using Cytoscape’s task manager, its progress is automatically tracked and reported by Cytoscape’s task monitor.\n\n\nResults\n\nThe GenomeSpace app enables Cytoscape to act as a tool in a workflow executed within the GenomeSpace web application. From the Cytoscape perspective, genomic data can come from numerous sources and can be rendered to numerous destinations, where the GenomeSpace file system can be a source, a destination, or both. To facilitate this, GenomeSpace also allows the user to identify a GenomeSpace-stored Cytoscape session file (.cys) while launching Cytoscape, thereby facilitating a seamless tool launch.\n\nAdditionally, the app adds menu items to Cytoscape to enable loading or saving a Cytoscape session, a network, or node or edge attributes in the GenomeSpace file system from within Cytoscape. Each menu item enables the user to navigate within the file system using a chooser.\n\nNote that the GenomeSpace app makes the GenomeSpace file system available to Cytoscape even without the user first executing the GenomeSpace web application. In this case, Cytoscape enables the user to log into GenomeSpace within Cytoscape, and then use the app-injected menus to access workflow-related data files.\n\nAs a demonstration of typical Cytoscape usage, we show how to use Cytoscape to integrate gene expression data with a pre-defined genomic network, where both the network and gene expression data reside in the GenomeSpace file system. The network was created in a prior Cytoscape session and stored in the GenomeSpace file system so it could be shared as a template with collaborators and integrated with various gene expression datasets produced during an ongoing study. The test network is a portion of the BioGrid H. sapiens network. The gene expression data represents the output of some previously executed GenomeSpace tool, such as Galaxy.\n\nTo load the template network into Cytoscape, use a web browser on a workstation having at least 6GB RAM. Log into Genomespace.org – you can easily create a user ID if you don’t have one. Right-click on the “Cytoscape 3” tool, and choose the Launch on File menu item. Choose the Cytoscape session file (.cys) containing the H. sapiens network template by browsing the file system to /bdemchak/F1000Example, then dragging it to the launch box and clicking Launch. Once Cytoscape has started and the network has loaded, you will see a small network in a view window (Figure 2).\n\nTo load a gene expression dataset, choose Cytoscape’s File \| Import \| Table \| GenomeSpace menu item, and use the chooser to select the enrichment data in the /bdemchak/F1000Example folder. Cytoscape will display an Import Columns from Table dialog – when you click on OK, note that the enrichment data (including values in the HSC1_1 column) has been added to the node table.\n\nFinally, to use the gene expression data to color the network, select Cytoscape’s Style tab and choose the RedYellowGreen style from the style dropdown. Nodes are colored by their associated HSC1_1 value, and nodes having no HSC1_1 value are left white (Figure 3). This style is part of the Cytoscape session file loaded for this example, and illustrates that all Cytoscape functionality is available in Cytoscape operating as a GenomeSpace tool.\n\nRefer to the Cytoscape manual (http://wiki.cytoscape.org/Cytoscape_3/UserManual) and tutorials (http://tutorials.cytoscape.org) for a more detailed treatment of Cytoscape workflows.\n\n\nConclusions\n\nThe GenomeSpace app shows how Cytoscape can be used as a plugin to a web application as part of a larger workflow, and then how to integrate external services into Cytoscape’s workflow.\n\nSoftware available from: http://apps.cytoscape.org/apps/genomespace\n\nLatest source code:\n\nhttps://github.com/idekerlab/genomespace-cytoscape-weblaunch\n\nhttps://github.com/idekerlab/genomespace-cytoscape\n\nSource code as at the time of publication: https://github.com/F1000Research/genomespace-cytoscape-weblaunch/releases/tag/v1.0\n\nhttps://github.com/F1000Research/genomespace-cytoscape/releases/tag/V1.0\n\nArchived source code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.1044112\n\nhttp://www.dx.doi.org/10.5281/zenodo.1053613\n\nLicense: Lesser GNU Public License v2.1", "appendix": "Author contributions\n\n\n\nTH, MS, TL and BD participated in the design of GenomeSpace app software. BD and TM authored this article. TI and JM provided funding and supervision.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe would like to acknowledge funding from National Human Genome Research Institute (P01 HG005062-03 assigned to Jill Mesirov) and Amazon Web Services. The GenomeSpace app was mainly authored at the University of California, San Diego, with the cooperation and assistance of the Broad Institute.\n\n\nAcknowledgements\n\nJohannes Ruscheinski authored early versions of the GenomeSpace app.\n\n\nSupplementary data\n\nTaskManager downloads GenomeSpace file\n\n\n\nSample JNLP file generated by cytoscape.php\n\n\n\n\nReferences\n\nWolstencroft K, Haines R, Fellows D, et al.: The Taverna workflow suite: designing and executing workflows of Web Services on the desktop, web or in the cloud. Nucleic Acids Res. 2013; 41(Web Server issue): W557–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltintas I, Wang J, Crawl D, et al.: Challenges and approaches for distributed workflowdriven analysis of large-scale biological data. Proceedings of the Workshop on Data analytics in the Cloud at EDBT/ICDT 2012 Conference, DanaC2012, 2012; 73–78. Publisher Full Text\n\nGoecks J, Nekrutenko A, Taylor J, et al.: Galaxy: a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences. Genome Biol. 2010; 11(8): R86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRocca-Serra P, Brandizi M, Maguire E, et al.: ISA software suite: supporting standards-compliant experimental annotation and enabling curation at the community level. Bioinformatics. 2010; 26(18): 2354–2356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReich M, Liefeld T, Gould J, et al.: GenePattern 2.0. Nat Genet. 2006; 38(5): 500–501. PubMed Abstract \| Publisher Full Text\n\nPerez-Llamas C, Lopez-Bigas N: Gitools: analysis and visualisation of genomic data using interactive heat-maps. PLoS One. 2011; 6(5): e19541. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu T, Ortiz JA, Taing L, et al.: Cistrome: an integrative platform for transcriptional regulation studies. Genome Biol. 2011; 12(8): R83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSynapse: In L. Sullivan (Ed.), The SAGE glossary of the social and behavioral sciences. (p. 504). Thousand Oaks, CA: SAGE Publications, Inc. 2009. Publisher Full Text\n\nThorvaldsdóttir H, Robinson JT, Miserov JP: Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform. 2013; 14(2): 178–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLotia S, Montojo J, Dong Y, et al.: Cytoscape App Store. Bioinformatics. 2013; 29(10): 1350–1351. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarinilli M: Java Deployment: with JNLP and WebStart, Sams, Indianapolis, IN 2001. Reference Source\n\nDemchak B, Hull T, Reich M, et al.: F1000Research/genomespace-cytoscape-weblaunch. ZENODO. 2014. Data Source\n\nDemchak B, Hull T, Reich M, et al.: F1000Research/genomespace-cytoscape. ZENODO. 2014. Data Source" }	[ { "id": "5291", "date": "30 Jul 2014", "name": "Aris Floratos", "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\nLanguage/spelling:The language is acceptable but would benefit from a more careful reading/editing. Some examples where improvements are possible are:“Javabased” should be corrected to “Java-based”. In the sentence that begins “Web browsers are defined to process a JNLP file URL by starting a specialized launcher…”, the portion that reads “are defined to” is extraneous. In the sentence that reads “Paradoxically, Javabased tools (such as Cytoscape) run directly on the user’s workstation instead of within a browser”, it is not clear what is paradoxical about the described execution protocol. The word “paradoxically” should/could be dropped or maybe be replaced by a more appropriate adverb.Comments on specific sections:Launch: there is a sentence that reads “GenomeSpace constructs the JNLP URL to contain a GenomeSpace file descriptor as a parameter, and the PHP script extracts it and defines it as a parameter to the LaunchHelper (see Supplementary Data).” It would add to clarity to explain what the file referenced is. GenomeSpace app: This section appears written with the assumption that the reader is already familiar with the Cytoscape app framework, essentially no background is provided. GenomeSpace communication strategy: Same comment as above: the article refers to Cytoscape-specific programming constructs without introducing them or providing relevant references for disambiguation. E.g., in the sentence “As with other apps, the app’s cyActivator initializes the app state, including gaining references to the standard Cy objects: application, network, view, and table managers”, terms like “cyActivator” and “Cy objects” are used without prior definition/explanation. The same issue exists with GenomeSpace-specific objects, e.g., the sentence “It also initializes basic CDK-related state (e.g., the GenomeSpaceContext root context)”. Demonstration: It would be helpful to describe what the columns of the input data file EnrichmentData.dat represent. Demonstration: The demonstration would benefit from a more detailed analysis workflow. E.g., a gene expression data set and a class file can be processed by GenePattern (or some other GenomeSpace tool) to generate a list of differentially expressed genes for some reasonable pair of case/control phenotypes and then feed the resulting file into Cytoscape. This could also help demonstrate the value proposition of GenomeSpace, i.e., providing the ability to seamlessly use multiple tools within an integrated environment. The authors may also want to briefly discuss why people are often interested in visualizing enriched/differentially-expressed genes in the context of a network, this would help the less-initiated better appreciate the value of having Cytoscape available as a GenomeSpace tool.Overall Impression:This is an informative article, describing how a specific tool (Cytoscape) has been integrated into an innovative framework (GenomeSpace) that enables seamless access to multiple bioinformatics tools. The main concern is overall readability in the absence of sufficient background on the Cytoscape app framework, to put the implementation of the GenomeSpace app in proper context. The other issues are minor and can be easily addressed.", "responses": [ { "c_id": "928", "date": "05 Aug 2014", "name": "Barry Demchak", "role": "Reader Comment", "response": "Thank you, Aris. Your observations and suggestions are very valuable, and we'll upgrade the paper accordingly. We really appreciate your detailed observations. Look for a new version in a couple of days." } ] }, { "id": "5651", "date": "04 Aug 2014", "name": "Gabriela Bindea", "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 “Cytoscape: the network visualization tool for GenomeSpace workflows” describes the GenomeSpace app, a tool that facilitates data analysis in Cytoscape for GenomeSpace users. This is a nice try to bridge two communities: cytoscapers & genomespacers.The resources and tools included in GenomeSpace are well presented, and the important data sharing possibility is underlined. At the same time, the extensive app collection of Cytoscape with its multiple analysis possibilities is very short mentioned. This could be maybe extended so that users of GenomeSpace could use Cytoscape not only to visualize their results obtained in GenomeSpace, but also to analyze their data with Cytoscape apps. But this will be probably the topic of another paper.For running the app, the computer should have enough memory, as mentioned in the article. Some GUI adjustments should be done for the different OS, for instance in Ubuntu the login dialog size should be increased, the buttons are not entirely visible at the moment. The new functionality is nicely underlined with icons, but is scattered throughout the menu. To increase the usability/visibility, after the login it would be good e.g. to add also the import/export buttons into the tool bar in top.The app performs well. It remains to the users of both communities to discover the multiple possibilities that this app offers.", "responses": [ { "c_id": "929", "date": "05 Aug 2014", "name": "Barry Demchak", "role": "Reader Comment", "response": "Thanks, Gabriela, for all of your helpful comments. The timing is good, as we're working a new Cytoscape release. We'll take a closer look at the dialog boxes. I don't think there's anything we can do about toolbar icons until the Cytoscape 3.3 release in 2015, but we'll certainly try to add the feature then." } ] }, { "id": "5449", "date": "04 Aug 2014", "name": "Jan Aerts", "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 paper \"Cytoscape: the network visualization tool for GenomeSpace workflows\", Demchak et al. describe the design, implementation and operation of an interaction layer between Cytoscape and GenomeSpace.Unfortunately, I was not able to run the example described in the manuscript, due to a \"unable to find the document\" error.Overall, the manuscript gives a fair description both of the technology used for integrating Cytoscape into GenomeSpace and its use. This however does mean that the actual focus is not completely clear. If the message is \"there is a new tool accessible from GenomeSpace\" which would be aimed at the end-user, then terms like \"JRAC\" and \"cyActivator\" cannot be expected to be understood without further explanation. Care should be given to explain these sufficiently enough for lay-people. On the other hand, the message could also be \"here is how to integrate a web-tool with a desktop application\", in which case the technical part might be described a bit more in depth.I expect this paper and the bridge between Cytoscape and GenomeSpace will be of considerable interest to the community.Minor textual remarks:In the abstract, there is an extra \"it\" in \"It has been downloaded it over 850 times\". Hyphenation has gone awry in many cases. Although this is not important for the contents of the manuscript, it does start to bother the reader after a while. These include (but are not limited to): \"GenomeS-pace\", \"submen-us\", \"grav-ity\", \"Genom-eSpace\", \"Genom-eSpaceContext\", \"attrib-utes\", and \"Bi-oGrid\".", "responses": [ { "c_id": "930", "date": "05 Aug 2014", "name": "Barry Demchak", "role": "Reader Comment", "response": "Thank you, Jan, for the thoughtful comments. We are looking into the issue of Cytoscape not launching from the web site. As a workaround, you can either hit the OK button on the browser error window, or launch Cytoscape on your PC before attempting to launch it from GenomeSpace. You can also get the Cytoscape session file directly from GenomeSpace by launching Cytoscape, and using the File \| Open from GenomeSpace menu item. Also, we'll address the mechanical and explanation issues in a revision in the next few days. Thank you!" } ] } ]	1	https://f1000research.com/articles/3-151
https://f1000research.com/articles/3-74/v1	18 Mar 14	{ "type": "Short Research Article", "title": "Role of bacteriophages in STEC infections: new implications for the design of prophylactic and treatment approaches", "authors": [ "Jaime H. Amorim", "Monica R. Jesus", "Wilson B. Luiz", "Bruna F.M.M. Porchia", "Rita C.C. Ferreira", "Esteban G. Servat", "Pablo D. Ghiringhelli", "Luis C.S Ferreira", "Leticia V. Bentancor", "Jaime H. Amorim", "Monica R. Jesus", "Wilson B. Luiz", "Bruna F.M.M. Porchia", "Rita C.C. Ferreira", "Esteban G. Servat", "Pablo D. Ghiringhelli", "Luis C.S Ferreira" ], "abstract": "Shiga toxin (Stx) is considered the main virulence factor in Shiga toxin-producing Escherichia coli (STEC) infections. Previously we reported the expression of biologically active Stx by eukaryotic cells in vitro and in vivo following transfection with plasmids encoding Stx under control of the native bacterial promoter. Since stx genes are present in the genome of lysogenic bacteriophages, here we evaluated the relevance of bacteriophages during STEC infection. We used the non-pathogenic E. coli K12 strain carrying a lysogenic 933W mutant bacteriophage in which the stx operon was replaced by a gene encoding the green fluorescent protein (GFP). Tracking GFP expression using an In Vivo Imaging System (IVIS), we detected fluorescence in liver, kidney, and intestine of mice infected with the recombinant E. coli strain after treatment with ciprofloxacin, which induces the lytic replication and release of bacteriophages. In addition, we showed that chitosan, a linear polysaccharide composed of D-glucosamine residues and with a number of commercial and biomedical uses, had strong anti-bacteriophage effects, as demonstrated in vitro and in vivo. These findings bring promising perspectives for the prevention and treatment of hemolytic uremic syndrome (HUS) cases.", "keywords": [ "DNA", "mice", "solutions", "eukaryotic cells" ], "content": "Introduction\n\nInfections by Shiga toxin-producing Escherichia coli (STEC) strains are a serious public health concern, resulting in diarrhea, hemorrhagic colitis, and haemolytic uremic syndrome (HUS).\n\nStx is the main virulence factor in STEC strains. The stx gene is present in the genome of prophages, which are similar to the bacteriophage lambda found in the lysogenic form of various E. coli strains. Previously we reported that the native promoter of the Stx-encoding gene can drive expression of the toxin in eukaryotic cells in both in vivo and in vitro conditions1,2.\n\nMany questions remain unanswered with regard to the mechanism by which STEC infection causes HUS. In particular, we are interested in understanding how Stx enters the systemic circulation and why only very small numbers of bacteria are sufficient to induce HUS in humans.\n\nBased on our previous observations that the native stx gene promoter is active in host cells, we seek to understand the role that bacteriophages play in the pathogenesis of STEC strains. Recently, it was reported that bacteriophages carrying the stx gene are required for the development of HUS in the murine model3. We hypothesise that eukaryotic host cells might be transduced with and/or infected by Stx-encoding bacteriophages, leading to Stx dissemination in vivo to enter the systemic circulation. This would also explain why very small numbers of bacteria are sufficient to develop HUS.\n\nIn order to test whether bacteriophages are responsible for the induction of HUS, we used an anti-bacteriophage agent to inactivate them. Chitosan, a linear polysaccharide polymer obtained after the deacetylation of chitin, the structural element in the exoskeleton of crustaceans, possesses strong antimicrobial activity against several pathogenic microorganisms4. Its antiviral activity was reported on the bacteriophage c2, which infects Lactococcus strains, and on bacteriophage MS2, which infects E. coli5 without affecting significantly the growth of the bacterial culture6. In order to test our hypothesis, which would make Stx-encoding bacteriophages a new target for preventing and treating STEC infections, we used chitosan as an anti-bacteriophage agent in vitro and in vivo.\n\nInactivation of bacteriophages was observed in vitro after incubation with chitosan, inhibiting both the infection of, and replication in bacterial cells, and the transduction of eukaryotic cells.\n\nGFP dissemination was significantly reduced in mice treated with chitosan following infection with a non-pathogenic strain carrying a bacteriophage in which the stx gene was replaced by the GFP-encoding sequence. Last, preliminary results showed partial protection by chitosan in vivo of mice infected with STEC.\n\nThese results contribute to understanding STEC infections, posing implications for a similar scenario to occur in other infections caused by bacteria carrying lysogenic bacteriophages.\n\n\nMaterials and methods\n\nC600ΔTOX:GFP, a lysogenized C600 strain carrying the 933W bacteriophage in which the stx gene was replaced by the gfp sequence, was generously provided by Dr. Alison Weiss7. EDL933W, an enterohemorrhagic E. coli (EHEC) strain carrying the wild-type bacteriophage from which C600ΔTOX:GFP was obtained, was generously provided by Dr. Luis Carlos de Souza Ferreira, LDV-USP, Brazil.\n\nBaby BHK-21 cells (Syrian hamster kidney fibroblasts from the American Type Culture Collection) cells were grown on 12-well plates (Nunc) in complete medium (10% fetal bovine serum in DMEM medium, Gibco, USA) for use in the transduction assay.\n\nC600ΔTOX:GFP was generously provided by Dr Alison Weiss7. This is a non-pathogenic phage resulting from purified 933W bacteriophage in which stx gene was replaced by gfp sequence (ϕ ΔTOX:GFP). Phages at a multiplicity of infection (M.O.I) equal to 1 were added to BHK-21 cells cultured the day before on 12 wells plate (Nunc). BHK-21 cells were counted with a Neubauer camera, and bacteriophage titer was measured by the titration assay as described below. Transduction of BHK-21 cells was enhanced by centrifugation at 1000 × g for 10 minutes at room temperature as previously reported1. After incubation at 37°C for 3 hours, the phage-containing medium was removed. Cells were washed twice with phosphate buffered saline (PBS) and then incubated in complete DMEM medium (Gibco, USA). Twenty four hours post-transduction, cells were washed with PBS, harvested and centrifuged at 2655 × g for 15 minutes. DNA was harvested from pellets after incubation for 5 minutes at 98°C in lysis solution (Tris pH8 50mM, SDS 2%, Triton-X100 5%) and the harvested DNA was used for PCR. Primers: Up-R 5′CCGCTCGAGACTAGTGCAAAAGCGAGCCTGGTAAATAAATATG3′; Up-D 5′GGAATTCCATATGCTCGTTGAGGCATATGAAAATCAGAC3′. The reaction was run in a Eppendorf Termocycler at an initial 92°C for 120 seconds and then at 92°C for 20 seconds and 60°C for 20 seconds and 72°C for 120 seconds for 35 cycles using primers giving a fragment of 1310 bp on the upstream region of gfp gene into the bacteriophage genome.\n\nThe C600ΔTOX:GFP strain was grown in Luria Broth (LB) plus 10 mM CaCl2 and chloramphenicol (Sigma) (15 μg/ml final concentration) overnight (ON) at 37°C under agitation. The ON culture was diluted to OD600nm = 0.1 in LB plus 10 mM CaCl2 and chloramphenicol (Sigma) (15 μg/ml final concentration). Induction was carried out by adding ciprofloxacin to a final concentration of 40 ng/ml8. Bacteria were incubated for 6 hours at 37°C under agitation. Cultures were then centrifuged at 5000 rpm for 15 minutes. The bacteriophage-containing supernatant was filtered with 0.2 μm filters and kept at 4°C until the titration assay was performed.\n\nE. coli strain Y1090 (ATCC 37197) was grown in LB plus ampicillin ON at 37°C under agitation. The ON culture was diluted 1:100 in LB plus ampicillin and incubated for 2 hours at 37°C under agitation. At the end of the incubation, 500 μl samples of E. coli Y1090 were incubated with 5, 50 and 100 μl of a suspension containing bacteriophages for 30 minutes at room temperature. At the end of this incubation, 3 ml of Top Agar (Tryptone 1%; NaCl 0.5%; Agar 0.7%) plus CaCl2 (10 mM final concentration) was added, and plated on LB-Amp agar plates. Plates were incubated at 37°C and lysis plaques were visually counted.\n\nϕΔTOX:GFP was incubated with 5 mg/ml of a chitosan (Sigma 448877) solution in phosphate buffer 10 mM, at pH = 7 for 10 minutes at room temperature, and bacteriophage titers were measured as described in titration assay section.\n\nChitosan was also used in the bacteriophage induction assay described above. Chitosan was added 2 and 4 hours post-induction and bacteriophage titers were analyzed at 6 hours post-induction.\n\nBALB/c mice were bred in-house at the animal facility of the Microbiology Department of the São Paulo University, Brazil. The experimental protocol of this study followed the ethical principles for animal experimentation adopted by the Brazilian College of Animal Experimentation (COBEA) and was approved by the Ethics Committee on Animal Experiments of the Institute of Biomedical Sciences (Protocol number 106), University of São Paulo, in accordance with the principles set forth in the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, 1985).\n\nMale mice aged 6 weeks (18 to 20 g) were used for the In Vivo Imaging System (IVIS). Immature male and female DBA-2 mice (17–21 days of age, approximately 8–11 g body weight) were used immediately after weaning for the infection assays with EDL933W strain (n = 4). Mice were maintained under a 12-h light-dark cycle at 22 ± 2°C and fed a standard diet and water ad libitum.\n\nImmature male and female DBA-2 mice (17–21 days of age, approximately 8–11 g body weight) were used immediately after weaning for the infection assays (n = 4).\n\nE. coli EDL933W (ATCC 43895) was used for infection of mice following the protocol previously reported by Brando and collaborators8. Briefly, E. coli EDL933W was grown in Tryptic Soy Broth (TSB, DIFCO, BD) ON at 37°C. The ON culture was centrifuged at 14000 rpm for 15 minutes and the bacterial pellets were washed twice in PBS. Pellets were resuspended to have a final concentration of 3 × 1012 CFU/100 μl per mouse.\n\nThe bacterial suspension was delivered directly into the stomach of mice after 8 hours of food starvation, via a 5-French paediatric feeding tube. After 4 hours of ingesting the bacterial suspension, mice were given food and water. Control animals received 100 μl of sterile PBS. Survival was observed for one week. Both groups were composed by 4 animals.\n\nTo analyze the effect of chitosan in vivo, immature male and female DBA-2 were infected as described previously and treated with 100 μl of a chitosan solution at a concentration of 5 mg/ml, orally administered 2 hours after infection. Survival was observed for one week.\n\nTwo-month old BALB/c mice were used to infect orally with C600ϕΔTOX:GFP. Bacteriophage induction in vivo was performed with ciprofloxacin as described immediately below. After 2 hours of induction with ciprofloxacin, 100 μl of chitosan solution at a concentration of 5 mg/ml was administered orally to the mice and GFP dissemination by IVIS was analyzed.\n\nThis time, we used two-month old BALB/c mice. An ON culture of C600:ϕΔTOX-GFP was used to infect them. The ON culture was centrifuged at 14000 rpm for 15 minutes at 4°C. The pellet was washed with PBS and centrifuged again at 14000 rpm for 15 minutes at 4°C. The pellet was resuspended in a solution of 20% sucrose to have a concentration of 1 × 109 CFU/mouse. Mice were inoculated orally with strain C600:ϕΔTOX-GFP and in vivo bacteriophage excision was induced following the procedures described by Zhang and collaborators8. The mice were sacrificed with CO2 inhalation 24 hours after bacterial inoculation. Blood, spleens, kidneys, lungs, brains, intestines, hearts and livers were harvested by surgical removal and kept in PBS solution and evaluated for GFP expression using the IVIS system. To determine the effects of chitosan in vivo, the mice received 100 μl of a chitosan solution at a concentration of 5 mg/ml.\n\nStatistical significance between treatments and controls was analyzed using the Prism 5.0 software (GraphPad Software), and the P value is indicated by asterisks in the figures.\n\nAll other data correspond to the means ± standard errors of the means (SEM) for individual mice. Statistical differences were determined using the one-way analysis of variance (ANOVA).\n\n\nResults\n\nBacteriophage lytic induction was triggered in E. coli C600ΔTOX:GFP using ciprofloxacin8. We observed a significant decrease in the optical density of the bacterial culture after addition of the antibiotic and the release of phages into the culture supernatant (Figure 1, panel A and B). The bacteriophage titer was analyzed at different time points and a significant increase was observed after induction (Figure 1, panel B). The effect of chitosan as an anti-bacteriophage agent was also examined. To this aim, we added chitosan at a final concentration of 5 mg/ml to the bacterial culture 2 or 4 hours post-induction, and we observed the complete inactivation of the ϕΔTOX:GFP, without measurable toxic effects to the bacterial strain (Figure 1, panels A and B).\n\nA. Growth curve: C600ΔTOX:GFP was induced with ciprofloxacin and the optical density was measured at 600 nm at 0, 2, 4 and 6 hours after induction. Non-induced C600ΔTOX:GFP was used as control. Chitosan was added at 2 or 4 hours after induction. B. Bacteriophage ϕΔTOX:GFP titer: bacteriophage titers were analyzed at 0, 2, 4 and 6 hours post-induction. Chitosan was added at 2 or 4 h post-induction. Asterisks represent P<0.05.\n\nWe previously reported the capacity of ϕΔTOX:GFP to transduce macrophages in vitro1. To further evaluate the ability of chitosan to inhibit bacteriophage transduction of mammalian cells, BHK cells were transduced for 3 hours with ϕΔTOX:GFP, ϕΔTOX:GFP plus chitosan or ϕΔTOX:GFP treated with DNAse. Addition of DNAse to the bacteriophage sample would preclude any free DNA in the bacterial lysates prior to the transduction of cells. Untreated cells were used as a control. As shown in Figure 2, the bacteriophage DNA was detected by PCR in mammalian cells, showing the capacity of the virus to transduce this cell line. However, when BHK cells were transduced with bacteriophages pre-incubated with chitosan, no phage DNA was detected, confirming the inactivating action of chitosan on bacteriophages. Bacteriophage DNA was also detected in cells transduced with ϕΔTOX:GFP treated with DNAse (Figure 2).\n\nA. PCR on DNA extracted from eukaryotic cells: 24 hours after transduction, BHK cells were washed and treated with Trypsin-EDTA solution. DNA was extracted and PCR was performed. Line 1: Cells transduced with ϕΔTOX:GFP. Line 2: Cells transduced with ϕΔTOX:GFP plus chitosan. Line 3: Cells transduced with ϕΔTOX:GFP previously treated with DNAse. Line 4. Untreated cells. Line 5. Positive control (ϕΔTOX:GFP DNA). Line 6. Negative control. Line 7. 1 kb ladder (Invitrogen).\n\nTo demonstrate the in vivo behavior of bacteriophages, mice were infected with the lysogenic E. coli C600ΔTOX:GFP strain, followed by oral administration of ciprofloxacin 1 hour or 2 hours later. In order to evaluate the effect of chitosan in vivo, a group of mice was administered with chitosan 2 hours post-induction and a control group of uninfected mice was evaluated for auto-fluorescence background control in each organ. Twenty four hours after infection, organs were harvested and examined using the IVIS. As shown in Figure 3, GFP was detected in the intestine, liver and, to a lesser extent, kidney of mice orally infected and treated with ciprofloxacin. Remarkably, the addition of chitosan 2 hours after infection caused a sharp decrease in GFP detection in organs of mice orally infected with the E. coli strain C600ΔTOX:GFP (Figure 3, panels A and B), indirectly indicating reduction of bacteriophages in the cells, GFP release and dissemination. Moreover, viable phages were detected via the lysis plaque assay in intestine homogenates and blood samples of infected mice, in which bacteriophages were induced by ciprofloxacin (data not shown).\n\nA. IVIS Representative image: ciprofloxacin was administered 2 hours post-infection to induce ϕΔTOX:GFP in vivo. A group of mice was treated with chitosan 2 hours after bacteriophage induction. All mice were sacrificed 24 hours post-infection and brains, hearts, lungs, livers, spleens, kidneys and intestines were harvested and analyzed by IVIS. Fluorescence intensity was recorded as photons/sec/cm2, and the signal intensity represents the amount of GFP present. B. Graphic of fluorescence intensity on GFP-positive organs. Four animals per group were analyzed and the fluorescence intensity was quantified using Living Imaging 4.3.1 in Calipter Life Sciences.\n\nIn order to evaluate the in vivo effect of chitosan during the infection process, mice were orally challenged with a wild-type EDL933W strain, based on the model described by Brando and collaborators9. Another mouse group was also treated with chitosan, administered orally 2 hours post-infection, and survival was followed for one week. In this preliminary study, partial protection was observed in mice treated with chitosan, resulting in a delay in the death time (Figure 4). Mice infected with EDL933W strain died at 72 hours post-infection, and mice infected followed by treatment with one dose of chitosan died at 168 hours after infection.\n\nMice were infected orally with EDL933W strain. Controls did not receive chitosan (dots and broken line) and the experimental group received chitosan 2 hours post-infection (square and fill line). Survival rates were observed for one week: two mice infected with EDL933W died 72 hours post-infection, while two mice infected with EDL933W plus chitosan died 168 hours post-infection. The remaining two mice of each group survived 192 hours post-infection.\n\n\nDiscussion\n\nLambda bacteriophages are used in gene transfer and vaccine delivery because of their capacity to transduce mammalian cells in vivo10. Tyler and collaborators recently showed that prophage induction is required for renal disease and lethality in the EHEC mouse model, suggesting that free bacteriophages encoding Stx may play a direct role in the disease3.\n\nIn previous reports, we have showed that the native phage promoter controlling Stx expression is active in eukaryotic cells as demonstrated both in vitro1 and in vivo2. Based on these results and the reports previously described, we sought to evaluate whether bacteriophages could be considered a target for treating STEC infections. To this aim, we measured GFP by the strain C600ΔTOX:GFP and the mortality of infected mice following bacteriophage induction, and in vivo inactivation upon chitosan treatment positive expression was analyzed. GFP was observed in liver, intestine and kidney by IVIS on mice in which the bacteriophage lytic phase was induced by ciprofloxacin following infection. Of particular relevance was the observation that chitosan exerted a direct inactivation effect on ϕΔTOX:GFP in vitro and drastically reduced the detection of fluorescence in mice orally infected with the C600ΔTOX:GFP strain. Bacteriophage transduction of mammalian cells was also inhibited after incubation with chitosan.\n\nOur findings indicate that chitosan possesses strong anti-bacteriophage properties in vitro and in vivo. This positively charged polymeric polysaccharide has been reported to inhibit other bacteriophages and probably acts through electrostatic interactions with negatively charged capsid proteins5. Based on these effects we propose that chitosan may be a viable alternative for the treatment of STEC infections. Chitosan is already used in food and medicine, and it is harmless to humans, making it a cheap and safe option for this application.\n\nThe fact that only partial protection was observed in vivo using chitosan may be due to its short half-life11. Our results may contribute to understand why only small numbers of bacteria are sufficient to induce HUS in humans. If bacteriophages are induced in the gastrointestinal tract, then replicate, infect bacteria in the intestine and transduce host cells, small numbers of bacteria should be enough to produce a Stx concentration sufficient to cause significant damage.\n\nAltogether, these findings suggest a paradigm change on the role of bacteriophages in STEC infections, indicating they may be responsible for the development of disease rather than their bacterial host. Thus, prophylaxis and treatment of human bacterial infections carrying virulence factors on lysogenic bacteriophages may require targeting of the bacteriophages instead of, or as well as, the bacteria and toxins involved.\n\n\nData availability\n\nFigshare: Data sets for bacteriophage induction and effect of chitosan. http://dx.doi. 10.6084/m9.figshare.96079412", "appendix": "Author contributions\n\n\n\nLVB designed and performed experiments, analyzed the data and wrote the manuscript; LCSF provided advice on experimental design, data interpretation, obtained funding and critical reading of the manuscript; JHA performed experiments and provided advice on experimental design, MJRR-WBL-BFMMP, PDG and RCCF provided advice on experimental design. EGS provided critical feed-back and editing on the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by PICT 2411 from the Agencia Nacional de Promoción Científica y Tecnológica, Argentina (to L.V.B) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazil (to L.C.S.F). LVB and PDG are members of the Research Career of CONICET (Consejo Nacional de Ciencia y Tecnología).\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 acknowledge Dr. Alison A. Weiss for providing strain E. coli C600: ΔTOX-GFP.\n\n\nReferences\n\nBentancor LV, Bilen MF, Mejías MP, et al.: Functional capacity of Shiga-toxin promoter sequences in eukaryotic cells. PLoS One. 2013; 8(2): e57128. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBentancor LV, Mejías MP, Pinto A, et al.: Promoter sequence of Shiga toxin 2 (Stx2) is recognized in vivo leading to production of biologically active Stx2. MBio. 2013; 4(5): e00501–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTyler JS, Beeri K, Reynolds JL, et al.: Prophage induction is enhanced and required for renal disease and lethality in an EHEC mouse model. PLoS Pathog. 2013; 9(3): e1003236. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong M, Chen XG, Xing K, et al.: Antimicrobial properties of chitosan and mode of action: a state of the art review. Int J Food Microbiol. 2010; 144(1): 51–63. PubMed Abstract \| Publisher Full Text\n\nLy-Chatain MH, Moussaoui S, Vera A, et al.: Antiviral effect of cationic compounds on bacteriophages. Front Microbiol. 2013; 4: 46. eCollection 2013. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKochkina ZM, Pospeshny G, Chirkov SN: [Inhibition by chitosan of productive infection of T-series bacteriophages in the Escherichia coli culture]. Mikrobiologiia. 1995; 64(2): 211–5. PubMed Abstract\n\nGamage SD, Strasser JE, Chalk CL, et al.: Nonpathogenic Escherichia coli can contribute to the production of Shiga toxin. Infect Immun. 2003; 71(6): 3107–3115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang X, McDaniel AD, Wolf LE, et al.: Quinolone antibiotics induce Shiga toxin-encoding bacteriophages, toxin production, and death in mice. J Infect Dis. 2000; 181(2): 664–70. PubMed Abstract \| Publisher Full Text\n\nBrando RJ, Miliwebsky E, Bentancor L, et al.: Renal damage and death in weaned mice after oral infection with Shiga toxin 2-producing Escherichia coli strains. Clin Exp Immunol. 2008; 153(2): 297–306. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLankes HA, Zanghi CN, Santos K, et al.: In vivo gene delivery and expression by bacteriophage lambda vectors. J Appl Microbiol. 2007; 102(5): 1337–49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoane RJ, Frier M, Perkins AC, et al.: Evaluation of the clearance characteristics of bioadhesive systems in humans. Int J Pharm. 1999; 178(1): 55–65. PubMed Abstract \| Publisher Full Text\n\nBentancor L, Ferreira LCS, Amorim JH: Data of bacteriophage induction and chitosan effect. Figshare. 2014. Data Source" }	[ { "id": "4170", "date": "31 Mar 2014", "name": "Mikael Skurnik", "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 by Amorim et al. deals with the role of bacteriophages in STEC-infections. The authors have earlier demonstrated that the stx genes can be expressed within eukaryotic cells, provided the prophage-carried stx-DNA is introduced there in naked form, i.e., in transfected plasmids. In the present work, the authors wanted to test/prove the hypothesis that the stx-carrying prophage upon induction in vivo could contribute to the toxin production. They also tested whether the polysaccharide chitosan has anti-stx phage effect. I have some major and minor points:MajorBacteriophage inactivation assay: The experimental design of the bacteriophage inactivation assay uses only one concentration of chitosan. To demonstrate specificity, dose dependence should be demonstrated. In addition, the in vivo dose of chitosan was not indicated in the methods section (100 µl/mouse of 5 ml/ml chitosan was given orally to mice as indicated in the Effect of chitosan in vivo section. The mouse experiments were performed with too small a number of mice. Figure 1A of growth curves is missing a crucial control. What happens to E. coli C600 under the ciprofloxacin treatment? Figure 1B: The lack of the 4 hr column in chitosan 4h post-induction does not seem logical to me. There should be a ca 6000 PFU/ml column similar to that one in the induced 4hr sample. This discrepancy should be explained. Figure 2: the PCR experiment does not provide evidence of transduction. The definition of transduction is that DNA moves from one cell to another. PCR detects the phage DNA either free in the cell cytoplasm or packed in endocytosed phage particles. Therefore, the authors need to demonstrate that infective phage particles disappear from infected cells. The experiment also does not exclude the possibility that phage particles are just adsorbed on the cell surface. The experiment reported in figure 3 should also include bacterial counts from the organs as it is very likely that live E. coli bacteria, after a massive dose of 1013 bacteria per mouse, end up in the organs. Therefore the authors should demonstrate that the GFP response is not from bacteria infected by the GFP-phages. The Figure 4 experiment was performed with only 4 mice. Such an experiment should not be shown in a publication. In addition, different chitosan doses should be tested here also.MinorIntroduction, paragraph 3: This statement on the low number of bacteria during infection should be backed up with a reference. Materials and Methods: Dr Alison Weiss is thanked twice for same strain. One time should be enough. In addition, in the acknowledgements she is thanked a third time. The bacterial strain designation in the latter is given differently than elsewhere in the text. Transduction of Eukaryotic Cells: C600ΔTOX:GFP is a bacterial strain, not a non-pathogenic phage. EHEC infection: The final concentration of CFU/100µl/mouse needs revision. Figure 2 legend: The path the sample takes in the gel is called the lane, not line.", "responses": [ { "c_id": "926", "date": "01 Aug 2014", "name": "Leticia Bentancor", "role": "Author Response", "response": "Major:A dose-response curve of chitosan was done. We used 5mg/ml, 2.5mg/ml and 1mg/ml of chitosan on purified bacteriophage. To evaluate it, the bacteriophage solution was incubated for 10 minutes at room temperature with the different doses of chitosan and the bacteriophage inactivation was evaluated with a lysis plaques assay. Chitosan at 1mg/ml lost the inactivation activity on the bacteriophages. Chitosan at 5mg/ml and 2.5 mg/ml showed a 100% efficiency on bacteriophage inactivation, however, 1mg/ml of chitosan showed a loss of inactivation, showing between 5-10% bacteriophage inactivation. This experiment was performed in triplicate.The effect of chitosan in vivo was evaluated using a final concentration of 5mg/ml of chitosan solution. Each mouse received 100ml, so, the dose used was 500mg/mouse.The Material and Methods section was changed as follows:“Immature male and female DBA-2 infected, as described previously, were treated with 100ml of a chitosan solution at a concentration of 5 mg/ml (500 mg of chitosan per mouse) was orally administrated 2 hours after infection and survival was observed.” The experiment was shown as a preliminary result and this work it is a short communication. The model used has some experimental problems for the ages of mice used. The experiment was started using 6 mice per group, but some mice died after inoculation and not for the infection. For this reason, we had shown only 4 mice per group. We repeat the experiment, and again we have the same problem, however, we can see the same partial effect of chitosan in vivo. To further analyze the effect observed, we will report the results on a new publication with more details. The controls we used were:Non-induced E. coli C600DTOX:GFP (as a negative control) in which we can observe the normal growth of bacteria without bacteriophage inductionInduced E. coli C600DTOX:GFP (as a positive control) in which we can observe how bacteriophage induction affect the growth of bacteria. If you are thinking about E. coli C600 in absence of lisogenic fDTOX:GFP, we do not have access to this strain. But, we think that the controls used are well done. As an observation, we can said that no significant change in the growth was observed on E. coli Y1090 used for bacteriophage titration assay.The observation is right; we made an error in the graph. The values were checked and the correct value was added to the new graph. We agree with the definition of transduction, it is the process by which DNA is transferred from one cell to another by a virus. In our previous paper, we used the same definition to evaluate the capacity of fDTOX:GFP to transduce macrophages. In this case, we observed GFP expression and we concluded that macrophages were transduced by fDTOX:GFP. In this report, we did a different approach and we use PCR to detect bacteriophage DNA inside the cell.After your opinion, we did two assays. First, bacteriophages inside the cell were analyzed for titration assay. Second, as a preliminary data, fStx2 was used to transduce Vero cells, as a representative Stx2-susceptible cell line. Infective bacteriophage particles were not detected on transduced cells. The assay was made using lysis plaque assay of cellular extracts.Results:In order to analyze the transduction by an additional method, Vero cells were transduced with fStx2 and cytotoxicity induced by Stx2 was evaluated by microscopy. Vero cells transduced with fStx2 (panel A) showed a similar cytotoxicity to that of cells incubated with 1 CD50 of purified Stx2 (panel B). No cytotoxic effects were observed in non-treated Vero cells (panel D). Vero cells transduced with a M.O.I. = 0,0625 did not shown cytotoxic effect, demonstrating the specificity of the effect observed by fStx2 (panel D). Materials and Methods: In vitro evaluation of the capacity of Bacteriophage 933W to transduce Vero cells.E. coli EDL933W (ATCC 43895) was used to purify fStx2. E. coli EDL933W strain was grown in Luria Broth (LB) overnight (ON) at 37°C under agitation. The ON culture was diluted to OD600nm = 0.1 in LB. Induction was carried out by adding ciprofloxacin to a final concentration of 40 ng/ml8 in main text. Bacteria were incubated for 6 hours at 37°C under agitation. Cultures were then centrifuged at 5000 rpm for 15 minutes. The bacteriophage-containing supernatant was filtered with 0.2 mm filters, precipitated and purified. Briefly, supernatant was incubated on ice with a PEG-8000/NaCl solution for 30 minutes. After that, the solution containing bacteriophages was centrifugated and washed. The pellet was resuspended in STE buffer (1ml of Tris pH8, 0,2ml of 0,5M EDTA, 2ml of 5M NaCl, water up to 100ml). Phages at a multiplicity of infection (M.O.I) equal to 1 were added to Vero cells. Transduction of Vero cells was enhanced by centrifugation at 1000 x g for 10 min at room temperature. After 24 hours post transduction, cells were examined by microscopy using Nikon Eclipse TE2000 (NIS-Elements imaging software) equipped with a CCD camera. Dilutions of fStx2 were made to demonstrate specificity. Vero cells were incubated with purified Stx2 as positive control. Additional figure. In vitro evaluation of the capacity of Bacteriophage 933W to transduce Vero cells. A. Vero cells transduced with fStx2 (M.O.I. = 1). B. Vero cells incubated with purified Stx2. C. Vero cells transduced with a M.O.I. = 0,0625. D. Vero cells with not treatment. E. coli C600DTOX:GFP is not an invasive bacteria. Also, E. coli O157:H7 is a non invasive strain; for this reason we do not check for bacteria in organs. Bacteria were checked only on lungs samples, just to see if the inoculation was right. Bacteria were not detected in lungs. The dose used was selected after a previous experiment in which we evaluated the sensibility of IVIS in our system. GFP is not the best fluorescent protein for IVIS system; so, we needed to use a high dose of bacteria. As we described in this work, bacteriophages were detected by lysis plaques assay in intestine homogenates and blood samples of infected mice. It is important to do a highlight in the case of intestine sample, as it is very difficult to find E. coli C600DTOX:GFP. First, because the huge amount of bacteria present in the sample, and also, because the bacteria lysis induced by bacteriophage excision. The experiment was shown as a preliminary result and this work it is a short communication. The model used has some experimental problems for the ages of mice used. The experiment was started using 6 mice per group, but some mice died after inoculation and not for the infection. For this reason, we had shown only 4 mice per group. We repeat the experiment, and again we have the same problem, however, we can see the same partial effect of chitosan in vivo. To further analyze the effect observed, we will report the results on a new publication with more details. Minor:The statement “…very small numbers of bacteria are sufficient to induce HUS in humans…” is taking the data published recently, in which the authors demonstrated that a concentration of Stx2 as low as 10 fM is able to induce ribosome damage and to modulate selected cell signaling pathways that change cellular functions. If 10 fM of Stx2 is enough, very small numbers of bacteria should be sufficient to induce HUS (Petruzziello-Pellegrini & Marsden, 2012). We deleted the sentence “…was generously provided by Dr. Alison Weiss” in the sub-section “Transduction of eukaryotic cells” of material and methods section. We changed “This is a non-pathogenic phage…” to “This is a non-pathogenic cell…”. However, this non-pathogenic cell produces the excision of a non-pathogenic phage.The dose is correct. We used a dose of 3 × 1012 CFU/mice in a volume of 100 ml. Line was changed for Lane." } ] }, { "id": "4510", "date": "17 Apr 2014", "name": "Raúl Raya", "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 written by Amorim et al. describes the anti-phage activity of chitosan on two variants (wild type and a derivative where the stx gene was replaced by the gfp gene) of the temperate Shiga-toxin producing phage EDL933W. The anti-phage activity was evaluated both in vivo and in vitro. The authors suggest that chitosan could be a viable alternative for the treatment of STEC infections. Major:Phage Induction/anti-phage effects of chitosan/Figure 1:It seems that chitosan not only sequesters free-phage particles, but also stimulates the growth of uninduced cells (see induced cells treated with chitosan 2 hours post-induction reached higher final OD values). So, does chitosan inhibit the induction process of the temperate phage? Or, does chitosan also adsorb/inactivate ciprofloxacin? Even though in the Materials and Methods a “Bacteriophage inactivation assay” is described, no data is presented. A dose-response curve should be presented, to determine the phage binding (inactivation) capacity of chitosan. In vivo experiments:If the authors suggest that Stx phages, rather than bacterial cells, may be responsible for the development of the STEC infections, why they did not use purified phage in the vivo experiments? Does chitosan adsorb/inactivate the Shiga-toxins? If so, may it explain the delayed response observed in Figure 4 (“EDL933W plus chitosan”)? Minors: Abstract:Provide a reference after “… plasmids encoding Stx under control of the native bacterial promoter.” Change “E. coli K12 strain” to “E. coli C600 strain”. Materials and Methods:Delete “… was generously provided by Dr. Luis Carlos de Souza Ferreira, LDV-USP, Brazil.”, since Dr. Ferreira is one of the authors of the manuscript. Check the sentence “…was generously provided by Dr. Alison Weiss”; it is repeated twice in the Materials and Methods, and also in the Acknowledgments. Change “This is a non-pathogenic phage…” to “This is a non-pathogenic cell…” Change “…complete DMEM medium” to “DMEM medium”. Or, if the “complete DMEM medium” contains 10% fetal bovine serum, change “…complete DMEM medium” for “…complete medium”. Please, indicate how DNA was harvested. Change “Tris pH8 50mM” to “Tris pH8 50 mM”. Change “BALB/c mice were bred…” to ““BALB/c and DBA-2 mice were bred…” Change “…under a 12-h light-dark…” to “…under a 12-hour light-dark…”. Delete the sentence “Two-month old BALB/c mice … and GFP dissemination by IVIS was analyzed”. It is redundant. Results:Figure 1B: Should “Bacteriophage/ml” be “PFU/ml”? Why phage titers are so low? Figure 1A: Change “hs” for “hours” or “h”. Figure 2: Lanes 5 and 6 should read: “positive PCR control” and “negative PCR control”, respectively. In lane 78, indicate in the figure the kb values of the ladder. Delete “viable” in “viable phages”. Were phages transduced or adsorbed to mammalian cells? The sentence “Mice were orally challenged with a wild-type EDL933W” is not correct, since there was a direct delivery of bacterial cells into the stomach of mice. Discussion: Check “deliv-ery”", "responses": [ { "c_id": "927", "date": "01 Aug 2014", "name": "Leticia Bentancor", "role": "Author Response", "response": "Phage Induction/anti-phage effects of chitosan/Figure 1:The higher final OD values determined on the cells treated with chitosan 2 hours post-induction, versus the OD value determined on un-induced cells, is not statistically significant. However, we tested if chitosan is capable of inactivating ciprofloxacin. Ciprofloxacin and chitosan were incubated for 10 minutes at room temperature with chitosan at 5mg/ml. After pre-incubation, the mix was used to induce bacteriophage excision. To see a more significant effect on bacteriophage excision, the induction was incubated overnight. The OD value measured showed a non-significant difference between non-induced culture and induced culture with the antibiotic pre-incubated with chitosan (the experiment was performed in triplicate).On the other hand, purified bacteriophages were incubated with chitosan, and bacteriophage inactivation was observed with a lysis plaques assay. Bacteriophages were incubated for 10 minutes at room temperature with chitosan at 5mg/ml. After incubation, bacteriophage inactivation was evaluated. The bacteriophage solution containing a titer of 4x103pfu/ml was 100% inactivated after chitosan incubation. The assay was performed in triplicate. This result showed the capacity of chitosan to inactivate bacteriophage in vitro.Also, we did a dose-response curve of chitosan. We used 5mg/ml, 2.5mg/ml, and 1mg/ml of chitosan on purified bacteriophage solution. To evaluate it, the bacteriophage solution was incubated for 10 minutes at room temperature with the different doses of Chitosan and the bacteriophage inactivation was evaluated with a lysis plaques assay. Chitosan at 1mg/ml lost the inactivation activity on the bacteriophages. Chitosan at 5mg/ml and 2.5 mg/ml showed a 100% efficiency on bacteriophage inactivation, however, 1mg/ml of chitosan showed a loss of inactivation, showing between 5-10% of bacteriophage inactivation. This experiment was performed in triplicate.In vivo experiments:Chitosan was analyzed in vitro and in vivo on fDTOX:GFP. Inactivation of fDTOX:GFP was observed in vitro with a lysis plaques assay. On the other hand, a decrease of GFP was observed in vivo by IVIS. These results shown that chitosan has the capacity to inactive bacteriophages in absence of Shiga-toxins. A direct action of chitosan on Shiga toxin was not evaluated in this work since we do not have purified Stx2 for such experiments. The authors are working on murine infection with fStx2, but the results obtained will be part of a new publication. Abstract:Provide a reference after “… plasmids encoding Stx under control of the native bacterial promoter.”The reference was provided. Change “E. coli K12 strain” to “E. coli C600 strain”.The change was made.Materials and Methods:Delete “… was generously provided by Dr. Luis Carlos de Souza Ferreira, LDV-USP, Brazil.”, since Dr. Ferreira is one of the authors of the manuscript.Answer: We deleted “… was generously provided by Dr. Luis Carlos de Souza Ferreira, LDV-USP, Brazil.” Check the sentence “…was generously provided by Dr. Alison Weiss”; it is repeated twice in the Materials and Methods, and also in the Acknowledgments. We deleted the sentence in the sub-section “Transduction of eukaryotic cells” of material and methods section. Change “This is a non-pathogenic phage…” to “This is a non-pathogenic cell…”We changed “This is a non-pathogenic phage…” to “This is a non-pathogenic cell…”. However, this non-pathogenic cell produces the excision of a non-pathogenic phage. Change “…complete DMEM medium” to “DMEM medium”. Or, if the “complete DMEM medium” contains 10% fetal bovine serum, change “…complete DMEM medium” for “…complete medium”.Complete DMEM medium contains 10% fetal bovine serum, so, we changed for the second option “…complete medium”. Please, indicate how DNA was harvested.Cells were harvested using Trypsin-EDTA solution. After that, DNA was harvested from pellets by incubation with lysis solution described in material and methods. We included “by Trypsin-EDTA incubation” to clarify the procedure. Change “Tris pH8 50mM” to “Tris pH8 50 mM”.The change was made. Change “BALB/c mice were bred…” to ““BALB/c and DBA-2 mice were bred…” We changed “BALB/c mice were bred…” for ““BALB/c and DBA-2 mice were bred…” Change “…under a 12-h light-dark…” to “…under a 12-hour light-dark…”.The change was made. Delete the sentence “Two-month old BALB/c mice … and GFP dissemination by IVIS was analyzed”. It is redundant.We have two different mouse models. First, we have the model used to analyze GFP dissemination in which we used two months old mice. Second, we have the model used to analyze protection effect in which we used immature mice. For this reason we clarify the model every time. Let me know if you consider that we need to delete the sentence “Two-month old BALB/c mice … and GFP dissemination by IVIS was analyzed”.Results:Figure 1B: Should “Bacteriophage/ml” be “PFU/ml”? Why phage titers are so low?Bacteriophage/ml was changed to PFU/ml as reviewer suggested. See below. It is true that bacteriophage titers are low. An optimization for bacteriophage purification was done to obtain a higher titer of bacteriophage. The antibiotics used to induce C600DTOX:GFP was selected as an alternative for mitomycin C. The efficiency of bacteriophage induction is strain dependent. Zhang and collaborators reported a titer equal to 1,3x105 pfu/ml using ciprofloxacin but they used pathogenic strain E. coli O157:H7. We also observed a higher titer inducing the EDL933W strain, for this reason we suppose that the low titer observed is dependent on the strain used. Figure 1A: Change “hs” for “hours” or “h”. The change was made. Figure 2: Lanes 5 and 6 should read: “positive PCR control” and “negative PCR control”, respectively. In lane 78, indicate in the figure the kb values of the ladder.The changes were made. Delete “viable” in “viable phages”. Were phages transduced or adsorbed to mammalian cells?We deleted “viable” in “viable phages”. In this context, phages purified from tissue were detected by lysis plaque assay. The sentence “Mice were orally challenged with a wild-type EDL933W” is not correct, since there was a direct delivery of bacterial cells into the stomach of mice. The sentence “Mice were orally challenged with a wild-type EDL933W” was change by “Mice were intragastrically infected with a wild-type EDL933W”.Discussion:Check “deliv-ery”We did not find deliv-ery in the Discussion section." } ] } ]	1	https://f1000research.com/articles/3-74
https://f1000research.com/articles/3-38/v1	06 Feb 14	{ "type": "Opinion Article", "title": "Ethical research standards in a world of big data", "authors": [ "Caitlin M. Rivers", "Bryan L. Lewis", "Bryan L. Lewis" ], "abstract": "In 2009 Ginsberg et al. reported using Google search query volume to estimate influenza activity in advance of traditional methodologies. It was a groundbreaking example of digital disease detection, and it still remains illustrative of the power of gathering data from the internet for important research. In recent years, the methodologies have been extended to include new topics and data sources; Twitter in particular has been used for surveillance of influenza-like-illnesses, political sentiments, and even behavioral risk factors like sentiments about childhood vaccination programs. As the research landscape continuously changes, the protection of human subjects in online research needs to keep pace. Here we propose a number of guidelines for ensuring that the work done by digital researchers is supported by ethical-use principles. Our proposed guidelines include: 1) Study designs using Twitter-derived data should be transparent and readily available to the public. 2) The context in which a tweet is sent should be respected by researchers. 3) All data that could be used to identify tweet authors, including geolocations, should be secured. 4) No information collected from Twitter should be used to procure more data about tweet authors from other sources. 5) Study designs that require data collection from a few individuals rather than aggregate analysis require Institutional Review Board (IRB) approval. 6) Researchers should adhere to a user’s attempt to control his or her data by respecting privacy settings. As researchers, we believe that a discourse within the research community is needed to ensure protection of research subjects. These guidelines are offered to help start this discourse and to lay the foundations for the ethical use of Twitter data.", "keywords": [ "The growing popularity of social media sites presents a unique opportunity to study human interactions and experiences. Twitter", "one of the most popular social media sites", "allows users to ‘microblog’ by sharing 140 character messages with their social network. Although Twitter doesn’t disclose the number of people who use its service", "estimates are in the hundreds of millions and perhaps as many as half a billion1. Approximately 340 million tweets are sent every day around the world2. Researchers have begun to use this data to answer questions in a variety of fields3–9. Recent reflections on the data collection practices of the US National Security Administration have spurred similar meditations on the ethics of digital research10. The concern is that Twitter data could conceivably be used in a way that violates the privacy and rights of the tweet authors." ], "content": "Introduction\n\nThe growing popularity of social media sites presents a unique opportunity to study human interactions and experiences. Twitter, one of the most popular social media sites, allows users to ‘microblog’ by sharing 140 character messages with their social network. Although Twitter doesn’t disclose the number of people who use its service, estimates are in the hundreds of millions and perhaps as many as half a billion1. Approximately 340 million tweets are sent every day around the world2. Researchers have begun to use this data to answer questions in a variety of fields3–9. Recent reflections on the data collection practices of the US National Security Administration have spurred similar meditations on the ethics of digital research10. The concern is that Twitter data could conceivably be used in a way that violates the privacy and rights of the tweet authors.\n\nTwitter data have already been used in a number of studies to detect influenza-like illness (ILI)3–6, risky behaviors associated with the transmission of HIV7, sentiments about childhood vaccination programs8, and political sentiments9. These study designs generally feature count data, rather than user-specific data. For example, there are multiple studies that compare the proportion of tweets about flu-related symptoms to public health data on influenza-like incidence (ILI). An increase in syndromic flu tweets might indicate that an outbreak is occurring. These data are usually reported either at the national level or without any geographic parameters. Another common study design aims to determine public sentiments by counting words, phrases, and emoticons that co-occur with keywords like ‘Obama’. These sentiment indicators can be used to infer public opinions about political elections, mental health, or consumer products.\n\nUnlike Facebook, Pinterest and other competitors, Twitter provides several application programming interfaces (APIs) that allow real-time access to vast amounts of content. Data streamed through the APIs include metadata about the authors, including the text location from their profile (e.g. ‘Baltimore’), their time zone, the time they sent the tweet, the number of friends and followers they have, the number of tweets they have ever sent, and more (Figure 1). Approximately 1% of tweets have a geolocation, which uses GPS to append the author’s precise geographic coordinates to the tweet. Geolocations are sufficiently detailed to determine from which wing of a building a tweet was sent. The default privacy settings do not enable geolocation, but do make a user's tweets and metadata available through the API. Users can modify their setting to make their profile private, which shields their account from public view online and from the API11.\n\nThere are numerous ways to access the data through these APIs: one method is through the ‘garden hose’, which is a random sample of approximately 1% of all live-streamed tweets. Other access methods include the search API which enables searching for particular users, hashtags, or locations, and author-specific queries which can retrospectively gather up to 3,200 tweets from a single user11. Furthermore, in 2010 Twitter donated its entire historical record of tweets to the US Library of Congress. Detailed plans for these data are not yet available, but the Library of Congress has indicated that it intends to collaborate with academic institutions to make the data available to researchers12.\n\nThe strength of tweets as a data source is in the volume; collection through the garden hose API brings in approximately 60,000–100,000 tweets per day. However because tweets are short and often lack context, it is difficult for computers to determine tweet content automatically. For this reason, researchers primarily use tweet data to conduct population-level research concerned with trends and patterns. Study designs rely on large volumes of data to accommodate false positives and negatives. A typical data set contains millions of tweets and many thousands of tweet authors. However, a user-centric use case involving Twitter is not inconceivable. Researchers interested in social network analysis, qualitative research, and rare-event topics may eventually turn to Twitter as a data source. Potential methodologies include building a social network out of @mentions (the @ is Twitter lexicon for referencing another user); mining qualitative data from specific user’s accounts; or conducting prospective research by following a person or small group of people over time. These user-centric approaches are fundamentally different from population-level studies, and may require different ethical considerations than aggregated study designs. Additional methodologies might also involve interacting with Twitter users, which will not be addressed here.\n\nUnder non-digital circumstances, ethics guidelines suggest that collecting information from a public space where people could ‘reasonably expect to be observed by strangers’ is considered appropriate even without informed consent13. According to these guidelines, tweets are text that users publish for the purpose of sharing with others. The weakness of this argument is that it fails to distinguish between population-level research and research focused on selected individuals. It would be clearly unethical for a researcher to follow one specific shopper around the mall and gather data exclusively about him without his consent. However, simply counting or observing behavior in aggregate at a mall is an acceptable research practice. The difference is that the latter example adheres to a level of privacy that the observed individual might expect from being in public, whereas the former violates those natural privacy boundaries. A similar distinction is needed in digital research.\n\nAs an example of the potential privacy pitfalls of digital research, suppose investigators were interested in the social networks of adolescents suffering from depression. A research plan might look like this: the investigators gather geocoded tweets that contain words relevant to the topic of interest, as shown in Figure 2. They filter for geocodes that correspond to school locations in order to identify adolescent users. From there, a simple query to the Twitter API returns a list of followers for each of those presumably depressed adolescents. They now have a social network. For each member of the network, they mine the user's tweet histories to find identifying details such as their real names. The researchers then use the gathered information to 'snowball' data collection by curating from a variety of different sources like Facebook, tumblr and the White Pages. They can collect birth dates, cell phone numbers, home addresses, favorite hangout spots, “likes” and “dislikes”, etc. The final result would be detailed demographic information for potentially thousands of people who exhibit symptoms of depression or are connected to a depressed adolescent. Current guidelines do not prohibit this kind of research activity. However, if the same information were collected through surveys or other traditional means, Institutional Review Board (IRB) approval would be needed.\n\nThe example tweet displayed is fabricated.\n\nAccording to the US Department of Health and Human Services Policy for Protection of Human Research Subjects, data that are publicly available are exempt from requiring IRB approval14. Because Twitter data are public, they technically fall under this exemption. Furthermore, Twitter’s privacy policy makes no secret of the fact that user data are indexed by search engines, archived within the US Library of Congress, and are available through an API15. However, it is unlikely that many users follow the link to read the lengthy and complex document. One study found that it would take 244 hours a year for an average internet user to read every privacy policy of the unique sites they visit16.\n\nThe US Consumer Privacy Bill of Rights (CPBR)17 may instead serve as a useful framework for guiding researchers conducting research with Twitter. The CPBR was issued by the Obama administration in February 2012 in order to “give consumers clear guidance on what they should expect from those who handle their personal information, and set expectations for companies that use personal data.” There are seven principles enumerated by CPBR in Table 1.\n\n\nProposed guidelines for the ethical use of Twitter data\n\nThis information should be published in manuscripts, published on the web for the public to access, and provided to IRB (when relevant). Going forward, collaboration between the research community and Twitter to provide information to users about ongoing research and relevant results may also be beneficial. Transparency regarding uses of Internet data for research purposes is needed for fostering ‘privacy literacy’ so that the users can make informed decisions about participating in Twitter.\n\nA tweet author discussing his mental health, for example, does not do so with the intention of sharing that data with researchers; he does it to communicate with his digital community. Qualitatively analyzing these communications as if they are offered for research consumption does not align with the context in which the tweets were created. Twitter participants can reasonably expect to rely on some anonymity of the crowd to manage privacy.\n\nTo preserve source anonymity, direct quotes or screen names are not publishable, nor are any details that could be used to identify a subject. Any and all information that could be entered into a search engine to trace back to a human source should be protected. A composite of multiple example tweets may instead be used for illustrative purpose. Geolocations in particular should be scaled to a larger geographic area in order to avoid violating the privacy of those tweet authors. The Title 13 of the Data Protection and Privacy Policy, the federal law under which the Census Bureau is regulated, expressly forbids publishing GPS coordinates18; researchers should adhere to this guideline as well.\n\nFocused collection is also important for preserving anonymity. It is possible to use data collected from Twitter to discern the identities of tweet authors, which can then be used to find and collect additional information from additional sources. For example an author’s username, identifying details provided in tweet texts, or geolocations could all be used to collect data about that individual from other sources like Facebook, LinkedIn, Flickr, or public records.\n\nResearchers may not follow a user on Twitter in order to gain access to a protected account. Doing so would violate that user’s efforts to control his or her personal data.\n\nThere is not currently an expectation that researchers engaging in research using Twitter will interface with their IRB. As discussed above, studies that could be conceived as individual-based should require IRB approval, whereas research designs that use data in aggregate (e.g. counts of keywords) may proceed without explicit consent. In turn, review boards should keep abreast of social network mining methodologies and corresponding ethical considerations in order provide informed guidance to researchers.\n\n\nConclusions\n\nResearch involving Twitter is growing in popularity, but the issues surrounding the ethics of using it as a data source have not yet been closely examined. There are hypothetical study designs that could use Twitter data in a way that violates the privacy and ethical treatment of participants. In order to avoid those misuses, six guidelines derived from the US Consumer Privacy Bill of Rights are proposed. We welcome discourse in the research community on this topic, and encourage further discussion.\n\nPlease use the #EthicalTwitter hashtag on Twitter to participate in this discussion online.", "appendix": "Author contributions\n\n\n\nCR conceived of the guidelines and drafted the manuscript. BL contributed to both the development of the guidelines and preparation of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nResearch reported in this publication was supported by the US National Institute of General Medical Sciences of the National Institutes of Health under award number 2U01GM070694-09. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Additional support provided by Defense Threat Reduction Agency Validation Grant HDTRA1-11-1-0016.\n\n\nAcknowledgements\n\nWe thank our external collaborators and members of the Network Dynamics and Simulation Science Laboratory (NDSSL) for their suggestions and comments.\n\n\nReferences\n\nTwitter Reaches Half a Billion Accounts, More than 140 Millions in the U.S. Semiocast. 2012. Reference Source\n\nTwitter turns six. Twitter Blog. 2012. Reference Source\n\nGinsberg J, Mohebbi M, Patel R, et al.: Detecting Influenza Epidemics Using Search Engine Query Data. Nature. 2009; 457(7232): 1012–1014. PubMed Abstract \| Publisher Full Text\n\nAchrekar H, Gandhe A, Lazarus R, et al.: Predicting Flu Trends using Twitter Data. The First International Workshop on Cyber-Physical Networking Systems. 2011; 713–718. Reference Source\n\nLampos V, Cristianini N: Tracking the flu pandemic by monitoring the social web. 2nd IAPR Workshop on Cognitive Information Processing. 2010; 411–416. Publisher Full Text\n\nSalathé M, Bengtsson L, Bodnar TJ, et al.: Digital Epidemiology. PLoS Comput Biol. 2012; 8(7): e1002616. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYoung S, Rivers C, Lewis B: Methods of using real-time social media technologies for detection and remote monitoring of HIV outcomes. Preventative Medicine. 2014.\n\nCampbell E, Salathé M: Complex Social Contagion Makes Networks More Vulnerable to Disease Outbreaks. Sci Rep. 2013; 3: 1905. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTumasjan A, Sprenger TO, Sandner PG, et al.: Predicting Elections with Twitter: What 140 Characters Reveal about Political Sentiment. Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media. 2012; 178–185. Reference Source\n\nOmand D: NSA leaks: how to make surveillance both ethical and effective. The Guardian. 2013. Reference Source\n\nTwitter Documentation. Twitter. 2013. Reference Source\n\nRaymond M: How Tweet It Is!: Library Acquires Entire Twitter Archive. Library of Congress Blog. 2010. Reference Source\n\nThe British Psychological Society. Report of the Working Party on Conducting Research on the Internet. Ethics. 2007; 23(4): 257–257. Reference Source\n\nCode of federal regulation: Protection of human subjects. Department of Health and Human Services. 2009. Reference Source\n\nTwitter Privacy Policy. Twitter. 2013. Reference Source\n\nMcDonald AM, Cranor LF: The Cost of Reading Privacy Policies. I/S: A Journal of Law and Policy for the Information Society. 2008. Reference Source\n\nConsumer Privacy Bill of Rights. 2012. Reference Source\n\nData Protection and Privacy Policy. U.S. Census Bureau. 2012. Reference Source" }	[ { "id": "3876", "date": "06 Mar 2014", "name": "Tristan Henderson", "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 opinion article is interesting and timely, as it is clear that the quantity of research using social media and other sources of personal data is increasing, and the potential concerns around this need to be considered by researchers and ethics committees. I found the current form of the paper to be rather unsatisfactory, however: it is not clear how or where the proposed guidelines would be applied, how they were developed, and most importantly there is no contextualisation of how this paper builds on the myriad discussions that have already taken place in this arena.The introductory section of the paper is fine, and provides a clear description of how Twitter works, which is no doubt of use for the intended audience. But the authors need to clearly describe the purpose of the guidelines. Indeed there needs to be some introductory text before the guidelines are introduced, as the current text is rather incongruent. Is the text on the Consumer Privacy Bill of Rights supposed to be in a separate section following the introduction?Moreover the aim and applicability of the guidelines is unclear. How would these guidelines be enforced? Would they be applied to all studies? Is there any mechanism for exceptions to this? For instance if I wanted to do a study comparing health behaviour on Twitter and Facebook, I would need to collect data from both networks. But this violates the guidelines. Or are these guidelines only for studies where IRB approval would not be obtained? There is much discussion over whether social media studies require any IRB approval at all 1,2,3. But your last guideline seems to disagree with this. The title of the paper refers to “big data” but the paper itself really concerns itself with Twitter (with some discussion of linkability). Are the guidelines to apply to all big data studies?Given that this paper really only targets US ethics committees and IRBs, I was surprised that there was no discussion of the Common Rule or the recent Menlo Report 4,5. An opinion article in Science last year also discussed US regulation 6.One of the guidelines explicitly describes the context in which a tweet was sent. But there is no mention of contextual integrity 7 which seems to be the most common framework for studying this, and has indeed been applied to social media research 8,9,10.Danah Boyd has talked widely about the ethics of big data research 11,12 and her work should be discussed. Concerns about Twitter specifically have also been raised 13,14 and Neuhaus and Webmoor specifically propose “agile ethics” for the study of such networks 15. How would agile ethics fit in with your guidelines?The implicit assumption in your guidelines is that studies should occur without consent and involve trawling the firehose. But these kinds of studies in themselves are controversial 16. Moreover is there any reason why informed consent could not be obtained for some of these studies, especially since you touch on consent when discussing your hypothetical shopping centre study? We have looked at this empirically 17.One striking difference between research and the collection of data for business purposes is that researchers are typically interested in reproducibility. Indeed many research funding bodies worldwide are now insisting on data archiving and sharing as a requirement, and this would apply to social media studies as well (something that has been criticised in the past 18). Should your guidelines not address data sharing? Technical challenges exist, including the linkability problems that you discuss 19 and the difficulty of anonymising data 20, while solutions are still nascent 21,22.The BPS reference (13) should be updated 23. Moreover you ought to compare and contrast other ethical guidelines for Internet research 24,25,26, and other proposals for social media and health research 27,28.At one point you mention “privacy literacy” but there is no reference and it was unclear what you mean here. Privacy “salience” has been discussed in the literature 29 but might be slightly different.Finally, I qualify my review by stating that I am not a life scientist and that this is the first review that I have written for F1000Research. Indeed, I had not heard of it prior to my review request. I do question the appropriateness of a scientific journal for this work. If the intention of the authors is, as stated, to encourage more discussion then it would be good to focus this in one place. Since F1000Research allows comments from readers, then perhaps this does provide a better avenue for discussion than over Twitter, which can be difficult over long time-scales and is not particularly persistent. But in my opinion an even better mechanism for discussion and debate would be to hold a workshop and then publish a workshop report. Certainly further discussion with other communities, and examination of the discussion that has already taken place, would be useful to inform and improve this document. In the first instance, I think it would be good to see a revised document that shows more awareness of the current work, and where controversy exists, discusses why the authors have taken their chosen stance when proposing their guidelines.", "responses": [] }, { "id": "5282", "date": "01 Jul 2014", "name": "Sherry Emery", "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 general, I think this is a very helpful piece, which lays out some practical and legitimate privacy risks that arise when working with social media data. And the authors' suggestions for the code of ethics are thoughtful, yet warrant some debate. My main concern is that corporate interests, most prominently Facebook - but also app developers and marketers by the legion - are using these data to both manipulate and profit from users digital networks and behaviors. More discussion of the appropriateness of applying traditional IRB standards to social media research is needed. In particular, the relative uselessness of the 'terms of use' in the context of IRB standards is notable. But in practice, this is the standard that commercial entities are applying. For researchers to be held to a different standard risks abdicating social media research to corporate interests--and the public good of generating knowledge from these data will be lost unless that knowledge directly serves somebody's bottom line. I'd like to see a more thorough consideration of the practical implications of potential privacy breaches from social media research. Would these meet the minimal risk criterion used by most IRBs? If so, then some of the suggestions in the proposed guidelines may be unnecessarily 'protective.'", "responses": [] } ]	1	https://f1000research.com/articles/3-38
https://f1000research.com/articles/3-201/v1	21 Aug 14	{ "type": "Method Article", "title": "Active transmembrane drug transport in microgravity: a validation study using an ABC transporter model", "authors": [ "Sergi Vaquer", "Elisabet Cuyàs", "Arnau Rabadán", "Albert González", "Felip Fenollosa", "Rafael de la Torre", "Elisabet Cuyàs", "Arnau Rabadán", "Albert González", "Felip Fenollosa", "Rafael de la Torre" ], "abstract": "Microgravity has been shown to influence the expression of ABC (ATP-Binding Cassette) transporters in bacteria, fungi and mammals, but also to modify the activity of certain cellular components with structural and functional similarities to ABC transporters. Changes in activity of ABC transporters could lead to important metabolic disorders and undesired pharmacological effects during spaceflights. However, no current means exist to study the functionality of these transporters in microgravity. To this end, a Vesicular Transport Assay® (Solvo Biotechnology, Hungary) was adapted to evaluate multi-drug resistance-associated protein 2 (MRP2) trans-membrane estradiol-17-β-glucuronide (E17βG) transport activity, when activated by adenosine-tri-phosphate (ATP) during parabolic flights. Simple diffusion, ATP-independent transport and benzbromarone inhibition were also evaluated. A high accuracy engineering system was designed to perform, monitor and synchronize all procedures. Samples were analysed using a validated high sensitivity drug detection protocol. Experiments were performed in microgravity during parabolic flights, and compared to 1g on ground results using identical equipment and procedures in all cases. Our results revealed that sufficient equipment accuracy and analytical sensitivity were reached to detect transport activity in both gravitational conditions. Additionally, transport activity levels of on ground samples were within commercial transport standards, proving the validity of the methods and equipment used. MRP2 net transport activity was significantly reduced in microgravity, so was signal detected in simple diffusion samples. Ultra-structural changes induced by gravitational stress upon vesicle membranes or transporters could explain the current results, although alternative explanations are possible. Further research is needed to provide a conclusive answer in this regard. Nevertheless, the present validated technology opens new and interesting research lines in biology and human physiology with the potential for significant benefits for both space and terrestrial medicine.", "keywords": [ "Critical medical situations can occur in space missions1. In such situations", "adequate treatment may require the use of drugs with potential severe side effects. Pharmacokinetics and pharmacodynamics can be modified under weightlessness conditions by many factors2", "3", "which may lead to undesired pharmacological effects. These parameters should be assessed carefully in order to ensure safe use of medications in space. However", "insufficient information exists in this regard in microgravity4", "5 and proper empirical evaluation is not feasible due to the significant associated risks." ], "content": "Introduction\n\nCritical medical situations can occur in space missions1. In such situations, adequate treatment may require the use of drugs with potential severe side effects. Pharmacokinetics and pharmacodynamics can be modified under weightlessness conditions by many factors2,3, which may lead to undesired pharmacological effects. These parameters should be assessed carefully in order to ensure safe use of medications in space. However, insufficient information exists in this regard in microgravity4,5 and proper empirical evaluation is not feasible due to the significant associated risks.\n\nABC (ATP-Binding Cassette) transporters are a large family of trans-membrane proteins widely present in many life forms, from bacteria to mammals. In humans, they play important roles in many physiological processes and in drug pharmacokinetics, pharmacodynamics, and drug-to-drug interactions of many currently used medications in oncology, emergency medicine, critical care and general practice6,7. Inhibition of these transporters has been correlated with clinically relevant metabolic disorders, drug overdosing, and drug toxicity7. On the other hand, their overexpression can lead to treatment failure due to enhanced drug cell clearance and increased drug excretion from the body7. Multi-drug resistance-associated protein 2 (MRP2) is a well-studied representative of the ABC transporters family, by which transport of glucuronide, sulphate and gluthatione-conjugated drugs is made against a concentration gradient requiring adenosine-tri-phosphate (ATP) hydrolysis8.\n\nGenetic expression of ABC transporters has been found to change in real and simulated microgravity conditions. Two independent studies evaluated gene expression pattern in two models of medically relevant species of microorganisms (Salmonella sp. and Candida sp.) and found upregulation of certain ABC transporters genes during short duration spaceflight missions9,10. Similarly, results from a microgravity simulation study showed that significant variation of certain ABC transporter levels could be found in the liver and kidneys of a murine model11. These changes could lead to a modified antimicrobial susceptibility in microgravity and to different pharmacokinetic characteristics of medications during spaceflight. Furthermore, recent reports suggest that microgravity can modify the activity of certain important cell components12–14, especially transmembrane proteins and ion channels15–17 with structural and functional similarities to ABC transporters. Changes in ABC transporters activity could lead to clinically significant metabolic disorders and potentially dangerous undesired pharmacological effects during spaceflight. However, there is no current technology that permits a functional evaluation of these cellular components in microgravity. Therefore, we developed a new approach combining the use of a currently available commercial assay, a high precision electromechanical system and an enhanced drug detection protocol to evaluate ABC transporters activity in microgravity conditions. A validation study of this new biotechnological approach is presented.\n\n\nGlossary\n\nMicrogravity: A condition in which there is very little net gravitational force, as of a free-falling object, an orbit, or interstellar space. In this text microgravity, weightlessness and zero gravity (0g) are considered equivalents.\n\nParabolic flight/parabolic flight campaign: To create a weightless environment, an especially prepared airplane flies in a long parabolic arc, in which two 2g periods are experienced while ascending (injection phase) and recovering (pull out phase) from the manoeuvre. In between these two periods, the aircraft enters into a free falling state, lasting approximately 22s, in which microgravity is obtained inside the cabin. A regular parabolic flight consists of 30 parabolas with a number of breaks. A parabolic flight campaign consists of three flights.\n\n\nMethods\n\nThis study was performed during the 51st European Space Agency (ESA) parabolic flight campaign, under the auspices of the “Fly your thesis” programme, and was developed following the specifications and experimental requirements of ESA/Novespace parabolic flight campaigns18.\n\nA modified, validated and commercial Vesicular Transport Assay® (Solvo Biotechnology, Hungary) was used to evaluate MRP2 transport activity of estradiol 17-β-glucuronide (E17βG) in microgravity during parabolic flights. A short overview of this assay is provided hereinafter and additional information can be found at the producer’s website (www.solvo.hu). Recombinant baculovirus infected insect cells (Sf9 cells) are used to produce microscopic vesicles containing at least 12–15% of the selected human transporter protein following a standardized and validated production protocol owned by the producer7. Vesicles are built allowing transport towards the interior of the vesicle. Once a reaction is triggered, the substrate (E17βG) is transported into vesicles and retained. Vesicles are filtrated using a fiberglass Milipore filter (< 1μm pore size) and eluted with pure methanol to retrieve substrate, which is later analysed. In the Vesicular Transport Assay®, a base sample suspension is produced containing selected vesicles (50 μg of pure MRP2 transporter per sample) and an assay mix composed by E17βG, co-transporters and ions. Concentrations of the assay mix and vesicle quantities followed the instructions contained in the protocol provided by the vesicle producer, however volumes were adjusted to adapt to minimum volume requirements of our automated electromechanical system and to reduce the error induced by mechanical actuators. By modifying the aforementioned base suspension, four different groups of samples were generated to evaluate four different assay conditions as recommended by the producer19. MRP2+ samples, containing a fully functional MRP2 transporter, were generated in order to evaluate the full transport capabilities of the transporter. NoATP samples were generated by using the same composition of MRP2+ samples but an ATP-free solution was injected instead at the moment of reaction start. These groups of samples permitted the evaluation of ATP-independent transport activity, known to be present in the ABC transporters family. Following vesicle producer recommendations, we also generated an additional group of samples to which benzbromarone 7.5 mM (BZM - specific non-competitive inhibitor) was added. In this set of samples, pharmacological inhibition capabilities in microgravity could be assessed. Finally, MRPdef negative control samples were generated from vesicles with mutated MRP-like transporters. These vesicles are provided by the producer and present no active transport capabilities. Therefore, they permit evaluation of simple diffusion of substrate across membranes. These samples were used as negative controls for transporter activity measurements (see supplementary material Appendix 1. Chemical products for a more detailed sample composition description). Confirmation of transporter activation requires the detection of net E17βG quantities in activated samples (MRP2+, noATP and BZM samples). Therefore, the signal from MRPdef samples was subtracted from the actual signal in each group of samples to obtain net transport activity. We performed a preliminary evaluation of the assay, which showed fast initial reaction speeds with most of E17βG being transported within few seconds. Samples were prepared in small syringes (Becton, Dickinson and Company, New Jersey, USA) containing 360 μl of the assay suspension (vesicles + assay mix), frozen immediately at -80°C and transported in dry ice to the parabolic flight campaign site. During the parabolic flight, the samples were kept in certified cold containers at 2–6°C before being used. The first five parabolas of each flight were excluded in order for the investigator to acclimatize to microgravity. The reaction was manually triggered 1s after the 6th parabola injection phase by a precise servo-mechanic injection of 250 μl 0.2M ATP mix or ATP-free buffer depending on the sample group (MRP2+, BZM and MRPdef samples vs. noATP samples). The reaction was stopped automatically inducing a drastic temperature reduction and substrate dilution by applying 1.5 ml of washing mix at 2–6°C after 19s, still in microgravity. This procedure was repeated identically for each parabolic manoeuvre. Washing mix composition followed standard protocol specifications by the vesicle producer but volumes were adjusted to adapt to minimum volume requirements of our automated electromechanical system and to reduce the error induced by mechanical actuators. E17βG-filled vesicles were recovered by Millipore fibreglass filtration. Once on ground, pure methanol elution was performed for E17βG recovery and stabilization before transport. Standard Escherichia coli glucuronidase enzymatic reaction was later required for glucuronide removal from estradiol. Final estradiol concentration was measured by gas chromatography coupled to a mass spectrometry detection system (6890-GC/5973-MSD, Agilent Technologies, California, USA) using a validated methodology20. Reference experiments were repeated on ground in 1g conditions using the same equipment, materials and procedures as those used during parabolic flights. The time lapse between sample stabilization with methanol and analysis was comparable in both cases. Sample analysis was performed using the same laboratory equipment and procedures. Results from 1g on ground reference experiments were compared with those obtained in microgravity conditions.\n\nIn order to evaluate vesicle microscopic structural integrity after gravitational stress, vesicles were exposed to 2g centrifugal forces in 60 consecutive 30s periods simulating a parabolic flight. Vesicles were compared to non-centrifuged controls at 200× under methylene blue tincture by experienced anatomo-pathologists (Pathology Department, Corporació Sanitària Universitària Parc Taulí. Sabadell, Spain). Qualitative analysis of shape and mean size estimation of vesicles was performed by experienced anathomo-pathologists, as well as a semi-quantitative vesicle number estimation using multiple 1 mm2 sector partial counting method.\n\nAn electro-mechanical engineering system was especially designed to undertake the required procedures in microgravity following the design and safety specifications for parabolic flights (Figure 1 & Figure 2). Four experimentation units composed the prototype, one per each specific group of samples. Accurate fluid displacement was achieved implementing high precision component manufacturing and accurate linear engines (Schneider Electric, France). These engines presented 0.05 mm of displacement error, which represented a calculated 16 μl fluid displacement error (< 10% in our experimental setup). Sample temperature was strictly controlled using a dedicated thermal control system, which consisted of four preheating chambers to raise the sample temperature from 2–6°C to 37°C and four reaction chambers to maintain the temperature at 37.3°C during the reaction. In order to protect the vesicle integrity, 45°C was established as the maximum permissible temperature at syringe wall in preheating chambers given syringe radius and its thermal conductivity. Several temperature probes feedback a central computer, which maintained the temperatures at the required levels (Minco, Minneapolis, USA; TC-Direct, Madrid, Spain). This system also enabled cold storage of ATP and washing mix at 2–6°C in two certified isothermal boxes (Tecnisample, Barcelona, Spain). All procedures were controlled by an electrical and control system, which provided electrical power, actuator control, synchronization, monitoring, logging and a user-friendly interface (Telemecanique, France) (see supplementary material Appendix 2. List of Components for a more detailed description of equipment and components used).\n\nThe designed electromechanical system was composed of three racks. In this figure a scheme of the main rack is presented, containing the four experimentation units. One of the four experimentation units is highlighted with a dashed line. Several electrical holes permitted connection to all electrical components in the main rack (sensors, actuators, linear engines, resistances). One ATP-WB hole permitted access to liquid conductions transporting ATP and Wash Buffer to each experimentation units. Lengths are expressed in mm.\n\nThis picture shows the final configuration of the experiment on board the parabolic aircraft. The three racks presented were, from left to right: storage rack, main rack, fluids and control rack. All principal electric components were placed in a sealed fire-proof cabinet underneath the main rack to protect the equipment from water contact in case of failure of the containment system.\n\nResults were analysed using SPSS v19.0 (International Business Machines, New York, United States of America). ANOVA test for multiple variable comparisons was performed to confirm differences in raw E17βG signal between the four sample groups in each gravitational condition. Bilateral Dunnett t post-hoc test was later used for comparing sample groups with MRPdef controls to confirm transporter activation. Provided equality of variances could not be assured and unbalanced size of groups being compared, non-parametric Mann-Witney U test was used to evaluate the differences between microgravity and 1g on ground samples. Statistical significance threshold was established at p < 0.05 bilaterally in all cases.\n\n\nResults\n\nAll four different assays presented expected E17βG signal profile in microgravity and on ground 1g conditions (Table 1), being MRP2+ samples the most active and MRPdef the less active group of samples in all gravitational conditions. The ANOVA test revealed statistically significant signal differences between sample groups in each gravitational condition (on ground F = 4.95, p = 0.005 and microgravity F = 13.45, p < 0.001) and post-hoc Dunnett t test confirmed statistically significant differences between MRP2+ and MRPdef in both microgravity and 1g conditions (Table 2). These results confirmed activation of the transporter in MRP2+ samples. MRP2 net transport activity of 1g on ground samples laid within commercial transport standards (4.61 ng in a range of 3.7 – 7.4 ng of E17βG in our experimental configuration), conversely microgravity samples suffered a 19-fold activity reduction with statistically significant differences of E17βG transported in all cases. The Mean activity differences between microgravity and 1g on ground samples were 10.3 ng of E17βG in MRP2+ samples (Mann-Witney U < 0.001; p < 0.001; CI 95% = 7.4 – 13.2 ng E17βG), 5.95 ng of E17βG in MRPdef samples (Mann-Witney U < 0.001; p < 0.001; CI 95% = 4 – 7.9 ng E17βG), 7.65 ng of E17βG in noATP samples (Mann-Witney U < 0.001; p < 0.001; CI 95% = 6.14 – 9.16 ng E17βG) and 9.12 ng of E17βG in BZM samples (Mann-Witney U < 0.001; p < 0.001; CI 95% = 7.9 – 10.3 ng E17βG).\n\nMRP2+: fully activated MRP2 transporter. MRPdef: MRP-like defective transporter. noATP: without ATP. BZM: Inhibition by benzbromarone. * Mean E17βG signal in nanograms. † Percentage of activity.\n\nMRP2+: fully activated MRP2 transporter. MRPdef: MRP-like defective transporter. noATP: without ATP. BZM: Inhibition by benzbromarone. CI 95%: confidence interval 95%\n\n* Mean difference is obtained by subtracting MRPdef signal from the signal detected in each group of samples. Net E17βG transported in nanograms.\n\nDenotes statistically significant\n\nThe use of consecutive 2g periods in order to simulate the effects of hypergravity present during a parabolic flight did not show differences in shape, size or number of vesicles between centrifuged vesicles and control samples. Aggregation of vesicles was not observed, neither macroscopically nor microscopically during this simulation. The thermal control system implemented in our prototype maintained the temperature of reaction chambers at 37.25°C (CI 95%: 37.14 – 37.34°C) during the whole duration of experiments and kept the sample syringe wall temperature below 45°C in preheating chambers in all cases.\n\n\nDiscussion\n\nTo our knowledge, this is the first validation study of a biotechnological approach aimed at evaluating human ABC transporters-mediated active transmembrane transport capabilities in microgravity. Despite the high signal variability expected, the precision attained by our electromechanical system and the enhanced sensibility of the detection protocol used, enabled us to detect net MRP2 transport activity in the short microgravity periods provided by parabolic flights. Furthermore, activation profile and signal levels detected in 1g on ground samples laid within expected commercial standards, confirming the validity of the equipment and procedures used.\n\nThe results obtained in microgravity samples showed a significant reduction in MRP2 transport activity, when compared to on ground 1g controls, but presented a comparable activation profile. In order to ensure reliability of these findings, vesicle stability during a parabolic flight was assessed on ground. Simulations revealed that vesicle microscopic structure could remain stable during parabolic manoeuvres, in spite of repeated hypergravity periods. Provided that materials, equipment, procedures and transportation means were equal for all samples in all gravitational conditions, we hypothesized that gravitational stress during parabolic flights may have exerted an effect upon the vesicle membrane or the transporter, preventing proper transport or retention of E17βG. However, such modification should occur at an ultra-structural level, out of reach of our microscopic evaluation. Unfortunately, given the lack of similar research precedents, an ultra-structural evaluation of samples was not planned and was not possible after analysis, since all microgravity samples had been eluted to retrieve E17βG. Therefore, the results of the present evaluation cannot provide a definitive explanation of the effect of microgravity upon ABC transporters-mediated active transmembrane transport. Notwithstanding the aforementioned limitation, we tried to elaborate a plausible hypothesis to explain the present observations.\n\nCompelling evidence indicates that gravity is able to affect cellular and even molecular interactions in numerous life forms, such as in certain protozoa12,14, bacteria15, plant cells13,16,17 or mammals21–23. Enzymes from the cell membrane24, cytoskeleton25, cytosol26 and nucleus24 have been found to be sensible to gravitational changes. In principle, gravity is a weaker force compared to intermolecular bonds, polar-apolar interactions and van der Waals forces, therefore it is not clear yet how gravity could affect enzymatic reactions. However, changes in polar-apolar interactions and conformational changes of lipid structures occurring in microgravity could explain some observations24,25. Structures such as ion channels and trans-membrane signalling systems can be significantly affected by changes in physicochemical properties of the lipid bi-layer cell membrane. Similarly, the mechanism of action of ABC transporters, although still not fully unveiled, is highly dependent on transporter-membrane interactions28, which play a principal role in transporter stabilization and substrate recognition. Interestingly, we were able to demonstrate a reduction of the signal detected in MRPdef controls, where only simple diffusion of substrate was possible. This finding could be explained by an impeded E17βG diffusion into vesicles, or by an increase in membrane permeability in microgravity, which would lead to increased E17βG leakage. In any case, the present observations are indicative of important structural changes in the lipid bi-layer in microgravity, which could be induced by modified polar-apolar interactions of lipid structures in microgravity, as suggested elsewhere28. In this context, the interaction of ABC transporters with the membrane, substrate recognition capabilities and transporter activation could become significantly affected with the potential for relevant biological, physiological and medical consequences in microgravity. However, additional research is required to confirm the present hypothesis.\n\nThere are many potential applications of the presented technology, which can benefit both space and terrestrial medicine. Insufficient information exists on drug effects in space4,5 and is often based in very reduced observational studies30 or in animal models31. While clinical experience throughout spaceflight history has shown no critical metabolic alterations in astronauts, few medications have been studied from a pharmacological point of view in space. A reduction of transmembrane drug transport capabilities in microgravity could lead to drug accumulation and potentially toxic effects during spaceflight. However an increase in membrane permeability would lead to an enhanced drug clearance, increased absorption and significant changes in pharmacological proprieties of drugs. Although not conclusive, our results warrant additional and thorough evaluation of this and other human drug transportation systems in a more stable microgravity platform. To this end, this study describes a new method, based on in vitro analysis of human drug transmembrane transport capabilities, by which an approximation to drug effects in microgravity can be made. Similarly, reduced activity of certain ABC transporters has been associated with a number of serious medical conditions on Earth. On the other hand, enhanced activity of ABC transporters can decrease intestinal drug absorption, facilitate hepatic and renal drug excretion, modify drug distribution, and limit drug penetration to certain body tissues. These transporters are known to be responsible for a significant portion of the variability in treatment response to certain drugs used in emergency and critical care medicine6–8 and can cause multidrug resistance in a number of solid and hematological neoplasias, where overexpression of certain ABC transporters is present6,7. Further research derived from this study may help elucidate the intricate mechanism of action of ABC transporters and provide new information for developing more effective treatments for oncology and other medical specialities based on microgravity effects upon these transporters.\n\nThere are several limitations in this study. First, we designed a novel electromechanical system to allow the required biochemical reactions to be undertaken precisely and automatically in parabolic flights. This equipment had never been tested before in such conditions. However, results from 1g on ground control samples laid within commercial transport standards, proving functionality and validity of the research equipment and protocols used. Additionally, MRP2 net E17βG transport activity in microgravity fell within the detection range of our enhanced detection method, which confirmed that the required sensitivity was reached. Second, as previously addressed by Macarrone et al.28, “microgravity is likely to favour the dispersion in water of less dense molecules such as lipids and the opposite is observed in hypergravity”. The contact surface available for ABC transporter exposition to drug-rich medium would be reduced, and transport capability would consequently be diminished, if vesicles precipitated during 2g phases in parabolic flights. Furthermore, vesicle structural integrity has never been evaluated in variable gravitational conditions. Nevertheless, no macroscopic precipitation of samples was observed in-flight due to combined effects of aircraft vibration and the ATP turbulent injection jet. On-ground hypergravity simulation data suggested that vesicles remained stable and did not aggregate after repeated 2g periods. However ultrastructural alterations affecting the membrane or the transporter could still be possible and were not evaluated. Further research should focus on performing a detailed ultra-structural evaluation of vesicles and transporters in a more stable microgravity platform and use whole human cells to provide a more conclusive answer on the potential physiological and clinical effects of our findings. Other factors were also considered as possible disruptors of results and were evaluated. The temperature was strictly maintained within a narrow margin during experiments but aircraft pressure was reduced in flight (800 hPa in flight). However, pressure variations were not expected to cause any modification in sample composition nor in transport activity in our experimental setup.\n\nIn conclusion, we validated a new methodology for evaluating ABC transporters activity in microgravity during parabolic flights. Our results demonstrated that despite time constrains, the combination of an adapted commercial assay, a highly accurate electromechanical system and an enhanced drug detection protocol can provide enough accuracy, reproducibility and sensitivity to detect transport activity in microgravity. A significant decrease in transport activity of fully activated samples and a reduction of signal detected in negative controls could be observed in microgravity. While the present results are insufficient for drawing conclusive explanations to these observations, we hypothesize that altered polar-apolar interactions induced by gravitational stress during parabolic manoeuvres affected vesicle’s lipid bi-layer membrane, eventually changing its physicochemical properties, limited simple diffusion phenomena and impaired transporter - membrane interaction. However, alternative explanations are possible. Therefore, additional studies will be required to assess membrane and transporter ultra-structures in microgravity, and to confirm these results. The novel methodological approach presented here opens new and interesting research lines in biology, microbiology and human physiology with the potential for significant benefits for both space and terrestrial medicine.\n\n\nData availability\n\nDataset 1. Data of active transmembrane drug transport in microgravity, 10.5256/f1000research.4909.d3416932", "appendix": "Author contributions\n\n\n\nSV conceived the study, performed analysis and interpretation of the data and drafted the manuscript. EC, designed biological experiments, collected data, contributed to analysis and interpretation of the data and to drafting of the manuscript. AR designed and built the electromechanical system, performed the collection of data, contributed to analysis and interpretation of the data and to drafting the manuscript. AG designed and built the electrical & control system, performed the collection of data, contributed to analysis and interpretation of the data. FF performed the collection of data and contributed to drafting of the manuscript. RT conceived the study, designed biological experiments, and contributed to interpretation of the data and to drafting of the manuscript. All authors 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\nSchneider Electric (Private entity), granted to Felip Fenollosa - Fundació CIM. Solvo Biotechnology (Private entity), granted to Felip Fenollosa - Fundació CIM. Fundació CIM (Public entity), granted to Felip Fenollosa - Fundació CIM. Institut Municipal d’Investigació Mèdica de Barcelona (IMIM) (Public entity), granted to Rafael de la Torre – IMIM. Lloveras (Private entity), granted to Felip Fenollosa - Fundació CIM. Tecnisample (Private entity), granted to Felip Fenollosa - Fundació CIM. Caixa d’Enginyers (Private entity) granted to Felip Fenollosa - Fundació CIM. Sinerges (Private entity), granted to Felip Fenollosa - Fundació CIM. Col•legi d’Enginyers Tècnics Industrials de Barcelona (Public entity), granted to Felip Fenollosa - Fundació CIM. Interempresas (Private entity), granted to Felip Fenollosa - Fundació CIM. Laboratorio para la Investigación en Espacio y Microgravedad (LEEM) (Public entity), granted to The ABCtr TEAM (SV, EC, AR, AG). Universitat Politècnica de Catalunya (UPC) (Public entity). Grant number: N/A. Granted to: Felip Fenollosa - Fundació CIM. Universitat Autònoma de Barcelona (UAB) (Public entity), granted to Rafael de la Torre – IMIM. European Space Agency via the Education Office and the “Fly your thesis” initiative, Grant number: 9140, granted to The ABCtr TEAM (SV, EC, AR, AG).\n\n\nAcknowledgements\n\nThe team of the “ABC transporters in microgravity” project would like to thank all sponsoring partners and supporting institutions, especially IMIM, Fundació CIM, Schneider Electric, Solvo Biotechnology, ESA Education Office and Novespace, and all individual collaborators who have worked in this project.\n\n\nAppendix\n\nFluid System\n\nBD: Becton, Dickinson and Company\n\nHeating System\n\nCooling System\n\nContainment System\n\nElectrical System\n\n\nReferences\n\nBillica RD, Simmons SC, Mathes KL, et al.: Perception of the medical risk of spaceflight. Aviat Space Environ Med. 1996; 67(5): 467–473. PubMed Abstract\n\nBarratt M, Pool S: Principles of Clinical Medicine for Space Flight. (Various chapters). ISBN: 9780387988429. 2008. Publisher Full Text\n\nBuckey JC: Space Physiology. Oxford University Press. (Various chapters). ISBN: 9780195137255. 2006; 304. Reference Source\n\nGandia P, Saivin S, Houin G: The influence of weightlessness on pharmacokinetics. Fundam Clin Pharmacol. 2005; 19(6): 625–636. PubMed Abstract \| Publisher Full Text\n\nGraebe A, Schuck EL, Derendorf H, et al.: Physiological, pharmacokinetic and pharmacodynamic changes in space. J Clin Pharmacol. 2004; 44(8): 837–853. PubMed Abstract \| Publisher Full Text\n\nBeringer PM, Slaughter RL: Transporters and their impact on drug disposition. Ann Parmacother. 2005; 39(6): 1097–1108. PubMed Abstract \| Publisher Full Text\n\nGalvinas H, Krajcsi P, Cserepes J, et al.: The role of ABC transporters in drug resistance, metabolism and toxicity. Curr Drug Deliv. 2004; 1(1): 27–42. PubMed Abstract \| Publisher Full Text\n\nvan A: Multidrug Resistance Protein 2. Transport Properties of a Drug Efflux Pump. Thesis. University of Nijmegen. ISBN 90–9013589–8. 1999. Reference Source\n\nWilson JW, Ott CM, Höner zu Bentrup K, et al.: Space flight alters bacterial gene expression and virulence and reveals a role for global regulator Hfq. Proc Natl Acad Sci U S A. 2007; 104(41): 16299–304. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrabbé A, Nielsen-Preiss SM, Woolley CM, et al.: Spaceflight enhances cell aggregation and random budding in Candida albicans. PLoS One. 2013; 8(12): e80677. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu Sk, Bai S, Javeri K, et al.: Altered cytochrome P450 and P-glycoprotein levels in rats during simulated weightlessness. Aviat Space Environ Med. 2002; 73(2): 112–8. PubMed Abstract\n\nBräucker R, Cogoli A, Hemmersbach R: Graviperception and Graviresponse at the Cellular Level. Astrobiology. The Quest for the Conditions of Life. Berlin Heidelberg Springer-Verlag. 2002; 287–333. Publisher Full Text\n\nChebli Y, Geitmann A: Gravity research on plants: use of single-cell experimental models. Front Plant Sci. 2011; 2: 56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHäder DP, Hemmersbach R, Lebert M: Gravity and the behaviour of unicellular organisms. Cambridge & New York (United Kingdom, USA): Cambridge University Press (Various chapters). 2005. Reference Source\n\nGoldermann M, Hanke W: Ion channel are sensitive to gravity changes. Microgravity Sci Technol. 2001; 13(1): 35–38. PubMed Abstract \| Publisher Full Text\n\nKordyum EL: Calcium signalling in plant cells in altered gravity. Adv Space Res. 2003; 32(8): 1621–1630. PubMed Abstract \| Publisher Full Text\n\nMorita MT: Directional gravity sensing in gravitropism. Annu Rev Plant Biol. 2010; 61: 705–20. PubMed Abstract \| Publisher Full Text\n\nNovespace A300 ZERO-G rules and guidelines. Novespace RG-2008–2. 2008.\n\nGlavinas H, Krajcsi P: Determination of the interaction of drugs with the MRP2 transporter using the estradiol 17–β-D-glucuronide vesicular transport assay (AP 122/2.0). Solvo Biotechnol. 2005.\n\nde la Torre R, de la Torre X, Alía C, et al.: Changes in androgenic steroid profile due to urine contamination by microorganisms: a prospective study in the context of doping control. Anal Biochem. 2001; 289(2): 116–23. PubMed Abstract \| Publisher Full Text\n\nAbraham S, Klein HP, Lin CY, et al.: The effects of space flight on some rat liver enzymes regulating carbohydrate and lipid metabolism. Adv Space Res. 1981; 1(14): 199–217. PubMed Abstract \| Publisher Full Text\n\nCogoli-Greuter M, Lovis P, Vadrucci S: Signal transduction in T cells: an overview. J Gravit Physiol. 2004; 11(2): 53–6. PubMed Abstract\n\nHughes-Fulford M: Function of the cytoskeleton in gravisensing during spaceflight. Adv Space Res. 2003; 32(8): 1585–1593. PubMed Abstract \| Publisher Full Text\n\nBoonyaratanakornkit JB, Cogoli A, Li CF, et al.: Key gravity-sensitive signaling pathways drive T cell activation. FASEB J. 2005; 19(4): 2020–2022. PubMed Abstract \| Publisher Full Text\n\nTash JS, Bracho GE: Microgravity alters protein phosphorylation changes during initiation of sea urchin sperm motility. FASEB J. 1999; 13(Suppl): S43–S54. PubMed Abstract\n\nMaccarrone M, Bari M, Battista N, et al.: The catalytic efficiency of soybean lipoxygenase-1 is enhanced at low gravity. Biophys Chem. 2001; 90(3): 303–306. PubMed Abstract \| Publisher Full Text\n\nGiachetti E, Ranaldi F, Vanni P: Enzyme catalysis in microgravity: an intricate problem to be solved. FEBS Lett. 2001; 504(1–2): 78–9. PubMed Abstract \| Publisher Full Text\n\nMaccarrone M, Finazzi-Agró A: Enzyme activity in microgravity: a problem of catalysis at the water-lipid interface? FEBS Lett. 2001; 504(1–2): 80. PubMed Abstract \| Publisher Full Text\n\nLoo TW, Clarke DM: Recent progress in understanding the mechanism of P-glycoprotein-mediated drug efflux. J Membr Biol. 2005; 206(3): 173–185. PubMed Abstract \| Publisher Full Text\n\nCintron NM, Putcha L, Vanderploeg JM: Inflight pharmacokinetics of acetaminophen in saliva. Results of the Life Sciences DSOs Conducted aboard the Space Shuttle. 1981–1986 (NASA). 1987; 27–42. Reference Source\n\nMerrill AH Jr, Hoel M, Wang E, et al.: Altered carbohydrate, lipid, and xenobiotic metabolism by liver from rats flown on Cosmos 1887. FASEB J. 1990; 4(1): 95–100. PubMed Abstract\n\nVaquer S, Cuyàs E, Rabadán A, et al.: Data of active transmembrane drug transport in microgravity. F1000Research. 2014. Data Source" }	[ { "id": "5908", "date": "28 Aug 2014", "name": "Thais Russomano", "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\nIt is a well written article and deserves to be indexed.My comments and suggestions are:Abstract should explain the method better, including more details of controls on Earth. It should contain some data (numerical data with the most important statistical analysis)Introduction is good.MethodsSamples and procedures: I suggest explaining the phases of the parabola and their duration. I also believe that it would be good to clarify how many parabolas were used for the tests.Figures are ok, although figure 2 does not add much to the study itself, unless some explanation is added to the figure and its caption.ResultsI believe that figures could show the microG changes betterIf figures are kept, please correct first line on Table 1 (data is misplaced)I think it should be 1G (not just in results but throughout the text), and then ground should be deleted (it is a pleonasm in this case)DiscussionIt is very good, clear, and consistentConclusionIt could be shortened a bit.It was very good to have considered and presented the limitations of the study.", "responses": [] }, { "id": "6930", "date": "05 Dec 2014", "name": "Steven Rees", "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 is an interesting contribution to the study of membrane transporter activity in microgravity. The authors discuss in detail the problems not addressed by the study, and important future studies for further validation. This is an excellent addition to the literature, and should be indexed.Introduction was a concise and accurate overview of ABC transporters and microgravity simulations. Good use of controls, including consideration of differential permeability of vesicles themselves due to microgravity No control with BZM/deformed MPR2 to see differences in noncompetitive inhibition (see if there’s activity in negative control vesicles (MRP2def) to see if relevant, particularly for the point about inherent changes in vesicle permeability mentioned in Discussion). Vesicle integrity was confirmed with 2g on-ground simulated flight campaign study (why not just run study at 1g alongside samples to eliminate confounding variables?). Also, is it apparent that the 2g on-ground study truly replicates the experimental flight campaign? Table 2: NoATP mean diference, microgravity, is 0.04, not 0.05 (according to Table 1). Same mistake with 1g on ground MRP2+, though this is most likely due to rounding. Following Table 2’s purpose, comparing relative difference (dividing microgravity and 1g mean differences by MRP2def) would have made data more comparable, given stark differences in means presented in Table 1. Data seem to have high variance, despite precision provided by the in-flight device. Comparison of various flight results in groups would hint at potential time-based variance or subtle differences in parabolic flight that could have contributed to these results (repeated flights, to achieve true replicates over pseudoreplicates, could have also aided in this, though may be impractical given predicted high cost of this single flight campaign). No data presented on vesicle integrity outside of discussion in Results; even if not significant, would have been useful. It is not clear that microgravity causes an increase in vesicle permeability given the quite reduced transport of E17BG in microgravity, as suggested in the discussion, though the ratio of E17BG transport in microgravity is higher than that of 1g transport. Future studies looking into this would be useful. More information on parabolic flight campaign would have been useful, not only for reproducibility but given the novelty of this procedure. Studying differential effects of microgravity on mammalian versus insect vesicles would be an important follow-up investigation, as suggested by the authors.", "responses": [] } ]	1	https://f1000research.com/articles/3-201
https://f1000research.com/articles/3-200/v1	20 Aug 14	{ "type": "Software Tool Article", "title": "High-Performance Mixed Models Based Genome-Wide Association Analysis with omicABEL software", "authors": [ "Diego Fabregat-Traver", "Sodbo Zh. Sharapov", "Caroline Hayward", "Igor Rudan", "Harry Campbell", "Yurii Aulchenko", "Paolo Bientinesi", "Diego Fabregat-Traver", "Sodbo Zh. Sharapov", "Caroline Hayward", "Igor Rudan", "Harry Campbell", "Paolo Bientinesi" ], "abstract": "To raise the power of genome-wide association studies (GWAS) and avoid false-positive results in structured populations, one can rely on mixed model based tests. When large samples are used, and when multiple traits are to be studied in the ’omics’ context, this approach becomes computationally challenging. Here we consider the problem of mixed-model based GWAS for arbitrary number of traits, and demonstrate that for the analysis of single-trait and multiple-trait scenarios different computational algorithms are optimal. We implement these optimal algorithms in a high-performance computing framework that uses state-of-the-art linear algebra kernels, incorporates optimizations, and avoids redundant computations,increasing throughput while reducing memory usage and energy consumption. We show that, compared to existing libraries, our algorithms and software achieve considerable speed-ups. The OmicABEL software described in this manuscript is available under the GNUGPL v. 3 license as part of the GenABEL project for statistical genomics at http: //www.genabel.org/packages/OmicABEL.", "keywords": [ "Current biomedical research is experiencing a large boost in the amount of data generated. In particular", "investigations involve human cohorts comprising hundreds of thousands of participants as part of nation-wide biobanking initiatives", "furthermore", "by using both arrays that include hundreds of thousands of single nucleotide polymorphisms (SNPs)", "and more recently", "exome and whole-genome re-sequencing", "the genomes of these participants are being characterized at an increasing level of detail", "bringing the number of features assessed to tens of millions. At the same time", "technologies for high-throughput assessment of different molecular “omics” phenotypes in large study cohorts are becoming more and more affordable. These molecular phenotypes characterize different classes and sub-classes of biological molecules", "their functional modifications and relationships. Examples include hundreds of thousands of epigenetic modifications (epigenome1)", "levels of tens of thousands of transcripts (transcriptome2", "3)", "metabolites (metabonome or metabolome4", "5)", "glycans (glyco(protein)ome6", "7)", "and proteins (proteome8). The evolution of current molecular techniques expands our capacity to access different components in the omics space", "and new prominent omics emerge (e.g. cellomics", "interactomics", "activomics). The study of the genetic control of different omics brings the promise of new fundamental and applied biological discoveries", "however", "such analyses pose “big data” challenges." ], "content": "Introduction\n\nCurrent biomedical research is experiencing a large boost in the amount of data generated. In particular, investigations involve human cohorts comprising hundreds of thousands of participants as part of nation-wide biobanking initiatives; furthermore, by using both arrays that include hundreds of thousands of single nucleotide polymorphisms (SNPs), and more recently, exome and whole-genome re-sequencing, the genomes of these participants are being characterized at an increasing level of detail, bringing the number of features assessed to tens of millions. At the same time, technologies for high-throughput assessment of different molecular “omics” phenotypes in large study cohorts are becoming more and more affordable. These molecular phenotypes characterize different classes and sub-classes of biological molecules, their functional modifications and relationships. Examples include hundreds of thousands of epigenetic modifications (epigenome1), levels of tens of thousands of transcripts (transcriptome2,3), metabolites (metabonome or metabolome4,5), glycans (glyco(protein)ome6,7), and proteins (proteome8). The evolution of current molecular techniques expands our capacity to access different components in the omics space, and new prominent omics emerge (e.g. cellomics, interactomics, activomics). The study of the genetic control of different omics brings the promise of new fundamental and applied biological discoveries; however, such analyses pose “big data” challenges.\n\nGenome-Wide Association Studies (GWAS) is an established tool for analyzing the genetic control of complex traits9. In GWAS, the association between millions of genetic markers (usually SNPs) and phenotype(s) of interest is studied, with significant associations highlighting the genomic regions harboring the functional variants involved in the control of the trait. While initially GWAS were mostly used to study common diseases, with the rising availability and affordability of omics phenotypes, this methodology is now also applied to investigate the omics space6,7,10–13, providing important insights into both the mechanisms underlying the genetic regulation of particular biological systems, and the determinants of human health and disease14–16.\n\nIn this work, we address the computational challenges posed by big-data GWAS. These challenges arise when size of sample under analysis is very large or when (potentially hundreds of) thousands of omics phenotypes are studied. We consider analyses facilitated by the use of linear mixed models (LMMs)17,18, which allow for modeling of correlations between phenotypes of relatives. The LMMs are among the most flexible and powerful methods to account for the genetic (sub)structure that inevitably occurs even in carefully designed large population-based studies. However, the increase in power and precision achieved through the use of mixed models comes with considerable costs in terms of computing time.\n\nRecent advances in GWAS using mixed models19–25 represent a breakthrough compared to older methods, and allow analyses of a limited number of traits in reasonably sized samples even on personal computers. Still, current algorithms and software may be prohibitively expensive for analysis of large samples when dealing with omics data, since the time needed for a multi-trait analysis is essentially that of a one-trait study multiplied by the number of traits. Under this scenario, the analysis of even relatively small samples sizes leads to extremely long wait times. Therefore at the moment, the LMM-based GWAS analysis of large cohorts (tens to hundreds of thousands of participants), and even small (thousands of participants) studies involving omics measurements, represents a considerable problem. This limitation compromises the analyses availability, the data-to-knowledge turnaround time, and leads to excessive energy spending.\n\nWith this work, we aimed to address the aforementioned problems for big-data LMM-based GWAS. To do so, we took advantage of properties specific to the LMM formulation of GWAS, and analyzed a number of possible algorithms applicable to the analysis of large data. By combining sophisticated linear algebra and optimization techniques, we produced a fast and scalable software. Our software facilitates GWAS of tens of thousands of samples and hundreds of thousands of omics phenotypes, without the need for super-computing facilities.\n\n\nMethods\n\nIn a nutshell, LMM models the phenotypes of a group of n studied individuals as a point in an n-dimensional space, which comes from a multivariate Normal distribution. The expected mean is modeled using a standard regression model as E[Y] = Xβ, where X is the design matrix which includes the genotypes of interest and other covariates, and β are fixed effects. The variance-covariance matrix is defined as M = σ2 · (h2Φ + (1 − h2)I); here, σ2 is the total variance of the trait, h2 is the heritability coefficient, I is the identity matrix, and Φ is the matrix containing the relationship coefficients for all pairs of studied individuals. GWAS are performed by consecutively including SNPs in the analysis model (usually one SNP at a time) and computing the association statistics for the included SNP, thus iteratively applying the model throughout the genome.\n\nThe statistical model considered in this work is the same as that outlined in previous works19–21, and proceeds with analysis in two steps: for each trait considered, we first estimate the matrix of (co)variances between phenotypes, and then we use it when estimating the SNP effects (see Supplementary Note S1 for mathematical details).\n\nFigure 1 illustrates how a multi-trait analysis consists of a series of t separate single-trait analyses, each of which, in turn, consists of a series of m Generalized Least-Squares (GLS) problems. The key to fast analysis algorithms is the realization that such problems are correlated, both along the m and the t direction; in big-data GWAS, any approach that ignores such correlations cannot be feasible in terms of time-to-solution.\n\nGWAS with multiple phenotypes requires the solution of m × t correlated Generalized Least-Squares (GLS) problems, originating a three-dimensional object B of size m × t × p. Along the t direction, the variance-covariance matrix M and the phenotype y vary, while the design matrix X does not; conversely, in the m direction, M and y are fixed while X varies. Specifically, X can be viewed as consisting of two parts, XL and XR, where the former is constant across the entire grid and the latter changes along m. The figure also captures GWAS with single phenotype, in which case the dimension t reduces to 1.\n\nAiming at supporting computational scientists in the design of efficient software, two of the authors recently developed CLAK, an algebraic system that replicates the reasoning of human experts for the automatic discovery of linear algebra solvers26. The core idea is to first decompose a target matrix-based problem in terms of library-supported kernels, and then apply algorithmic and algebraic optimizations. Since the decomposition is not unique, CLAK returns not one but a family of possible solutions, all mathematically equivalent, but exhibiting different space and time complexity.\n\nWith the help of CLAK, we generated many solutions to perform the aforementioned GWAS analyses (for a representative list, see Table S2). Each solution was subjected to the analysis of its computational complexity, expressing the cost in terms of the number of samples, markers, and traits in question. Interestingly, depending on the number of traits, the best theoretical performance was attained by two different solutions, which we named CLAK-Chol and CLAK-Eig (described in Table S1), respectively. Figure 2 shows the surfaces representing the time complexity of CLAK-Chol and CLAK-Eig as a function of the number of traits and markers analyzed; the solid curve denotes the crossover between the surfaces. When fewer than four traits are considered, CLAK-Chol attains better theoretical performance; on the contrary, for a higher number of traits, CLAK-Eig is expected to perform better (see also Table 1, and Supplementary Methods).\n\nThe brown and blue surfaces indicate the number of operations performed by CLAK-Eig and CLAK-Chol, respectively, for a given number of SNPs and traits. The crossover curve suggests that for analyses with more than just a handful of traits, CLAK-Eig is the fastest algorithm.\n\nThe variables n, m and t denote the sample size, the number of genetic markers, and the number of traits, respectively. v is the average number of iterations necessary to estimate the model parameters σ2 and h2 (see “Time complexity” in Supplementary Note S1).\n\nThe idea underlying CLAK-Chol is to linearly transform the input data to de-correlate the observations. To this end, the variance-covariance matrix M is formed explicitly, and its triangular Cholesky factor L is computed (LLT = M); through this factor, each SNP X and each trait y is then linearly transformed, giving raise to a sequence of Ordinary Least Squares problems of the form b := (XTX)-1XTy. While such problems are solvable with standard techniques, CLAK-Chol takes advantage of the fact that the covariates are fixed for all SNPs, hence lowering the computational complexity. In single-trait analyses, it is also possible to exploit the fact that the matrix M is symmetric and positive definite. Although asymptotically the Cholesky factorization is equivalent to an eigen-decomposition, in practice it requires 10 times fewer operations. Moreover, instead of using the eigenvectors (a full matrix) to rotate the SNPs, the Cholesky factor (a triangular matrix) allows us to transform them at half the cost. For more details, see the work we have previously published27 describing the CLAK-Chol algorithm.\n\nThe design of CLAK-Eig is based on three insights: 1) For a given study, the relationship matrix Φ is constant across both SNPs and traits; 2) since the variance-covariance matrix M is built by merely shifting and scaling relationship matrix, its eigenvectors are the same as those of Φ, and its eigenvalues are obtained by shifting and scaling those of Φ; 3) the inverse of M is easily expressed by inverting the diagonal matrix containing its eigenvalues (note that in our solutions we never explicitly invert matrices; we instead factor them, and operate with their factors). Together, these insights suggest that the eigen-decomposition of Φ can be computed once and for all, and most importantly, the eigenvectors of the relationship matrix can be used to rotate all the SNPs and traits only once. After the data is rotated, the computation of the mixed model based GWAS can be carried out by means of a grid of inexpensive Weighted Least Squares problems.\n\nOnce the initial eigen-decomposition of Φ is available, the complexity of CLAK-Eig is determined by three operations: the rotation of the SNPs, the rotation of the traits, and the solution of the Weighted Least Squares problems. The dominant term depends on the size of population (n), number of SNPs (m), and number of traits (t). When n > t (or n > m), the overall time complexity comes from the rotation of the SNPs (or the traits), and amounts to O(n2m) (or O(n2t)); if instead both t and m are larger than n, then the dominant term comes from the Least Squares problems, and is linear in population, SNPs and traits: O(nmt) (Table 1). Note that the CLAK-Eig algorithm is a generalization of the eigen-decomposition based algorithms published before (e.g.22,28) for a case of multiple trait analysis.\n\nCompared with current state-of-the-art algorithms19,21,22 in multi-trait analyses, CLAK-Eig achieves a lower computational complexity. As shown in Table 1, there are two scenarios of interest, depending on whether the number of traits is larger than the population size or not. In the first case (t > n), which is probably the most typical for current and near-future omics studies, the time complexity of CLAK-Eig is linear on the number of markers, traits, and samples; by contrast, all the other methods have quadratic complexity with the sample size. In the second case (n > t), which takes place for smaller ‘omics’ and also will become more common with the increasing affordability of omics technologies and hence larger sample sizes, the cost of CLAK-Eig is determined by the sample size and the number of SNPs, and its complexity is a factor t lower than other methods.\n\nFor both CLAK-Eig and CLAK-Chol, the space complexity is mainly determined by the square of the sample size; also, a minimum of one trait, one SNP, and the p covariates must reside in memory. In total, our methods only require enough memory to accommodate n2+(2+p)n entries. If multiple SNPs and/or traits fit in main memory at once, —e.g., dozens or hundreds of them—the computational throughput of our methods improves noticeably. In this case, the space requirement becomes n2+(k+p)n, where k is the number of SNPs and traits resident in memory. As examples, for sample sizes of 10,000, 20,000 and 40,000, the n2 space requirement translates to 1, 3, and 12 GBs, respectively. More details on space complexity are provided in Supplementary Note S1.\n\n\nResults\n\nTo demonstrate the practical advantages of CLAK-Eig and CLAK-Chol, we implemented these algorithms in the OmicABEL software package. In doing so, we tailored our implementations to save intermediate results across adjacent problems; we also re-organized the calculations to fully benefit from both the efficiency of highly optimized linear algebra kernels, and the parallelism offered by modern computing platforms.\n\nSince the size of the datasets involved in GWAS is considerably larger than the memory capacity of current processors, input and output data can only be stored in disk devices. Aware that the penalty for accessing information residing on disk is enormous—several orders of magnitude greater than the cost for performing one arithmetic operation—it is imperative to handle these big-data efficiently. By means of asynchronous transfers between memory and disk, our algorithms achieve a perfect overlap of computation and data movement. As long as the relationship matrix fits in the main memory, and regardless of the size of the data sets—both in terms of SNPs and phenotypes—, the processor never idles waiting for a transfer to complete, thus computing at maximum efficiency.\n\nWe compared the GWAS run-time of CLAK-Chol and CLAK-Eig as implemented in OmicABEL with that of several well-established packages: EMMAX19, FaST-LMM22 (two-step approximation), and GWFGLS (implementation of the mmscore method of ProbABEL21 in the MixABEL-package29). In the experiments, we considered three different scenarios, varying one among sample size n, number of SNPs m, and number of traits t, while keeping the other two values constant (Figure 3). A description of the experimental setup is provided in Supplementary Note S2 and Supplementary Note S3.\n\nPanels (a) and (b) include timings for EMMAX, GWFGLS, FaST-LMM, and our OmicABEL software, for single trait analyses; (c) and (d) present a comparison of EMMAX, GWFGLS, FaST-LMM, and our OmicABEL software, in the case of multiple traits. In (a), the number of SNPs is fixed to m = 107 and the sample size n ranges from 1,000 to 40,000. In panel (b), the sample size is fixed to n = 10,000 and the number of SNPs m ranges between 106 and 3.6 × 107. In (c) and (d), n = 1,000, m = 106 and t ranges from 1 to 100,000.\n\nIn the first scenario (single trait and m = 10,000,000 SNPs; the sample size varies from 1,000 to 40,000), all methods exhibit a quadratic behavior, and CLAK-Chol is the only algorithm that completed all tests within 25 hours (Figure 3(a)). For the largest problem considered (sample size n = 40,000), the speed-up over the next-fastest software (FaST-LMM) is 8.3: 205 vs. 25 hours; when n = 1,000, the speed-ups over GWFGLS, FaST-LMM and EMMAX are 24, 32 and 68, respectively: 86, 112 and 240 vs. 3.5 minutes.\n\nThe second scenario (single trait and sample size of 10,000; the number of markers varies between 1 and 36 millions) shows a linear dependence on the number of genetic markers for all software packages. Again, CLAK-Chol attains the best timings, outperforming FaST-LMM, GWFGLS and EMMAX by a factor of 11.7, 93.6 and 298, respectively (Figure 3(b)), when the number of analysis SNPs is 36 millions.\n\nFinally, the third scenario illustrates the analysis of multiple phenotypes (sample size of 1,000 and 1,000,000 genetic markers; number of traits varies between 1 and 100,000). The (estimated) time required for these analyses is presented in Figure 3(d). Note that the time scale on this graph is in years. Due to CLAK-Eig’s linear time complexity with respect to sample size, SNPs and traits, its advantage becomes most apparent: when thousands of traits are considered, CLAK-Eig outperforms GWFGLS, FaST-LMM and EMMAX by a factor of 1012, 1352, and 2789, respectively (Figure 3(d)), bringing the execution time down from months to hours.\n\nWe applied the OmicABEL (CLAK-Eig) to study 107,144 metabolomic traits in a sample of 781 people from a genetically isolated population of the Vis island (Croatia). These data are part of the EUROSPAN data set reported in the works of 11, and 12. In short, the data comprise plasma levels of 23 sphingomyelins (SPM), 9 ceramides (CER), 56 phosphatidylcholines (PC), 15 lysophosphatidylcholines (LPC), 27 phosphatidylethanolamines (PE), and 19 PE-based plasmalogens (PLPE). From these data, additional traits were defined by aggregating species into groups with similar characteristics (e.g. unsaturated ceramides), and also by expressing data as molar percentages (instead of absolute concentrations) within classes. Following the standards accepted in genetic analysis of metabolomics data30, in this work, 328 such measurements served as a base to compute all pair-wise ratios, resulting in 107,584 traits, which were analyzed for association with 266,878 SNPs. More details about the data are provided in Supplementary Note S4.\n\nPreviously, it took several weeks to accomplish the original analysis of only few hundreds “original” traits. However, using our OmicABEL software and a computer with 40 cores, we were able to finish the analysis of more than 100,000 traits in only 8 hours.\n\nTable S4 shows the results for five SNPs previously reported11 to be significantly associated with levels of circulating sphingolipid concentrations. The best results obtained for these SNPs when using the original and derived traits are reported. The implicated traits were also analyzed by using other approaches: the full likelihood ratio test based LMM18,22 (as implemented in MixABEL::FMM), Grammar-γ25 (as implemented in the GenABEL-package29), and the two-step approach19–21 (as implemented in MixABEL::GWFGLS). From Table S4 one can see that our results are consistent with those obtained by other methods. Additionally, Table S3 summarizes the results concerning heritabilities of analyzed traits, while Table S5 lists the Geinflation factor λ obtained when analyzing the selected traits with different methods. More details of the analysis of this human metabolomics data set are provided in Supplementary Note S4.\n\n\nDiscussion\n\nIn contemporary human genomics, methods and tools face the additional challenge posed by the sheer size of the datasets. Big data are produced from the investigation of large human cohorts including hundreds of thousands of participants; these massive samples facilitate the identification of small effects and lead to important biological insights. Large data also come from the field of functional (gen)omics, which aims at establishing the functional roles of genetic variants; hence GWAS are increasingly applied not only to study complex traits in large cohorts, but also to understand the regulation of human and animal transcriptome15,31,32, metabolome10–12, glyco(protein)ome6,7 and other types of omics data. Results of these studies are used to uncover the link between these molecular phenotypes and high-level complex traits, including human diseases14–16.\n\nIn recent years, linear mixed model was accepted as a powerful tool for whole-genome analysis of genetic associations17,18. Most current LMM-based methods for GWAS19–24 exhibit linear dependency of the compute time on the number of genetic polymorphisms and traits studied, but at least quadratic dependency on the sample size. A notable exception from the latter “at-least-quadratic” rule is the GRAMMAR-Gamma method25 and a method based on low rank approximation of the similarity matrix22 - with the latter exploiting the ideas similar to EIGENSTRAT approach33 and the methods assuming the adjustment of the model for top Principal Components of the kinship matrix variation. However, the computational advantage of these methods comes at the cost of mathematical approximation. For example, the GRAMMAR-Gamma method, while extremely fast, and showing excellent results for human studies, is less suited for analyses of samples with uneven genetic structure; adjusting for top principle components (and EIGENSTRAT) is known to provide incomplete correction for stratification in case of complex kinship. Increasing sample sizes and availability of molecular omics phenotypes lead to “big data challenges” and the computational throughput of LMM’s starts being more and more of an issue, which at present sometimes cannot be resolved without resorting to supercomputing facilities.\n\nWith this work, we address the problem of mixed-model based whole-genome analysis of genetic association for an arbitrary number of traits. We describe the CLAK-Chol and CLAK-Eig algorithms and software tools (OmicABEL package) to address LMM-based GWAS. Specifically, our CLAK-Chol will be useful for investigation of complex traits in very large (tens of thousands of individuals) samples, while CLAK-Eig will be a useful tool for the investigation of genetic control of different omics, potentially including hundreds of thousands or even millions of features.\n\nAs for our CLAK-Chol approach, we are not aware of similar, Cholesky-based solutions proposed before. The CLAK-Eig approach behind our solutions bear similarities, and actually reduces to previously suggested methods (e.g.22,28) when the number of traits is one. It is also worth mentioning the Matrix eQTL software34, which, while not implementing the LMM, in many respects exploits the problem-specific properties of multi-trait GWAS in the ways similar to ours.\n\nThe key achievement of this research is that it facilitates big-data LMM-based GWAS without supercomputers. For a sample problem with a population of 1,000, three covariates, one million SNPs and 100,000 traits, we estimate that the available methods would require the entire Sequoia supercomputer (equipped with 1.5 million cores)1 for about 3 minutes; by contrast, using a common 40-core compute node (see Supplementary Methods, Note S2), our method completes within a day and reduces the energy consumption by a factor of 200 (estimated). It should be noted that this impressive speed-up comes at the price of additional assumption of complete data. For many types of omics assays assumption of absence of missing data could be (almost) true, and a small proportion of missing data could be imputed (in the simplest case – replaced with average value) with little negative effect onto statistical properties of the method. However, for the omics assays which produce large proportion of missing data, our CLAK-Eig method in its current formulation and implementation would be inapplicable, unless the missing values could be reliably imputed.\n\nIn case of single-trait analysis, our results are somewhat less impressive, and our CLAK-Chol solution outperforms advanced current methods (e.g. FaST-LMM) by about one order of magnitude for large-sample-size problems. It is worth mentioning that the latter speed-up becomes possible because we show that for single-trait GWAS problems our CLAK-Chol algorithm is superior to CLAK-Eig, but other current methods are actually implementing solutions similar to our CLAK-Eig algorithm to address the single-trait GWAS problem.\n\nFurther optimizations of our solutions are possible, for instance by exploiting the structure of the kinship matrix. A “compressed MLM” approach was proposed for decreasing the effective sample size of datasets by clustering individuals into groups20; similarly, the fast decaying and possibly sparse structure of the kinship matrix can be exploited to lower the number of mathematical operations. Caution must be exercised in the interpretation of findings resulting from GWAS analyses as they may generate false positives if the multiple testing problem is not addressed adequately. A conservative strategy to determine whether an association is statistically significant would be to apply a Bonferronni correction, that is, in our example analysis of 107,144 traits, the conventional genome-wide significance threshold p-value of 5 · 10-8 should be replaced by 4.7 · 10-13. This is the common approach applied in the metabolomics studies (see10,30). On the other hand, this threshold would probably be too conservative, given that many of the measurements may be highly correlated. Several methods have been introduced recently35–37, which may help to overcome this problem; however, this topic lays outside the scope of the current work.\n\nFor the analysis of specific omics data, our methods (and software) might require some modifications. For example, in the genetic regulation of human transcriptome, large attention is dedicated to cis-eQTLs, computationally a relatively simple task. In contrast, our implementation is tailored to perform full GWAS for every trait analyzed. While OmicABEL could be used for the identification of trans-eQTLs, one should be aware of the specifics of the analysis of this type of data (e.g. allele-specific expression in RNAseq studies) and a body of methods developed (e.g. methods to account for influences of hidden factors38,39).\n\nWe foresee that the primary use of our algorithms and software is within the domain of analysis of complex traits in very large samples and for the genetic analysis of “omics” data. However, potentially, there are other uses. The same set of methods and tools can be used for scanning through other omics in e.g. search for biomarkers for a complex trait or in order to determine functional relations between different omics. For example, one may be interested in doing epigenome-wide scans relating the epigenome to a complex trait (or other omics, such as metabolome). Under this use, the genomic inputs would be replaced by epigenomic data. Advanced statistical and machine learning methods, such as penalized regression, can make use of joint analysis of up to several hundreds of thousands of predictors40. One of the common scenarios include the construction of millions of features, and their filtering for further joint analysis—a task which can be also effectively addressed by our methods. Finally, our algorithms can be easily extended, e.g. to search for interactions.\n\n\nConclusions\n\nWe demonstrated that different computational algorithms are optimal for the problems of single- and multi-trait Mixed-Model based GWAS, and implemented these algorithms in a freely available OmicABEL software.\n\n\nSoftware availability\n\nThe OmicABEL software implementing computational methods described in this manuscript is available as part of the GenABEL project for statistical genomics at http://www.genabel.org/packages/OmicABEL. The web-page provides the link to OmicABEL tutorial, giving examples of its use.\n\nhttps://r-forge.r-project.org/scm/viewvc.php/pkg/OmicABEL/?root=genabel\n\nhttp://dx.doi.org/10.5281/zenodo.1099941\n\nGNU GPL v. 3", "appendix": "Author contributions\n\n\n\nYA and PB jointly designed and supervised the project. DFT and PB developed the methods and algorithms. DFT designed the software, and, together with SZS and YA, analyzed the data. CH, IR and HC provided the data for demonstration analysis. DFT, YA and PB wrote the original version of the paper. All authors contributed to review of the manuscript during its preparation and agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nPB and DFT gratefully acknowledge the support received from the Deutsche Forschungsgemeinschaft (German Research Association) through grant GSC 111. The VIS study in the Croatian island of Vis was supported through the Core grant from the Medical Research Council UK and a grant of the Ministry of Science, Education, and Sport of the Republic of Croatia (number 108-1080315-0302). EUROSPAN (European Special Populations Research Network) was supported by the European Commission FP6 STRP grant number 018947 (LSHG-CT-2006-01947). The work of SZhS was funded by a joint grant from RFBR and the Helmholtz society (Joint Research Groups, 12-04-91322). The work of YA was supported by grant from the Russian Science Foundation (RSCF, grant no. 14-14-00313). The work of IR, CH and HC has received funding from the European Union’s Seventh Framework Programme (FP7-Health) under the grant agreement no. 305280 (MIMOmics).\n\n\nAcknowledgements\n\nWe are grateful to Tatiana Axenovich, Felix Agakov, Xia Shen, and Lars Rönnegård for fruitful discussion of the manuscript, and to Nadezhda Belonogova for help with setting up the analysis. We thank the Center for Computing and Communication at RWTH Aachen for the computing resources.\n\n\nSupplementary Tables\n\nThe label “Original” refers to the average heritability of the original traits (intercept of the model); similarly, “cer-cer” refers to the heritability of ratios involving CER, “cer-lpc” to the heritability of ratios involving CER and LPC, and so on.\n\nLead trait: the original trait for which the association signal was most significant in Vis data. Lead percentage: the molar percentage, for which the most significant association was obtained. Lead Ratio: the ratio, for which the most significant association was obtained. OmicA: OmicABEL, our implementation of CLAK-Eig; FASTA: the two-step approach19–21 (as implemented in MixABEL::GWFGLS); LRT: the full likelihood ratio test based LMM18,22 (as implemented in MixABEL::FMM); Gra-γ: Grammar-γ25 (as implemented in the GenABEL-package29).\n\n\nNotes\n\n1As of January of 2014, Sequoia is the third fastest supercomputer in the world: http://www.top500.org/lists/2013/11/.\n\n2Recall that our algorithms do not require large amounts of available RAM, as long as it accommodates the kinship matrix, the algorithms will complete.\n\n\nReferences\n\nFlintoft L: Human epigenomics: Putting epigenetic variation on the map. Nat Rev Genet. 2009; 10: 663–663. Publisher Full Text\n\nde Koning DJ, Haley CS: Genetical genomics in humans and model organisms. Trends Genet. 2005; 21(7): 377–381. PubMed Abstract \| Publisher 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\nNicholson JK, Lindon JC, Holmes E: ‘Metabonomics’: understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological nmr spectroscopic data. Xenobiotica. 1999; 29(11): 1181–1189. PubMed Abstract \| Publisher Full Text\n\nRaamsdonk LM, Teusink B, Broadhurst D, et al.: A functional genomics strategy that uses metabolome data to reveal the phenotype of silent mutations. Nat Biotechnol. 2001; 19(1): 45–50. PubMed Abstract \| Publisher Full Text\n\nLauc G, Essafi A, Huffman JE, et al.: Genomics meets glycomics: the first GWAS study of human N-glycome identifies HNF1α as a master regulator of plasma protein fucosylation. PLoS Genet. 2010; 6: e1001256. Publisher Full Text\n\nLauc G, Huffman JE, Pučić M, et al.: Loci associated with N-glycosylation of human immunoglobulin g show pleiotropy with autoimmune diseases and haematological cancers. PLoS Genet. 2013; 9(1): e1003225. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltelaar AF, Munoz J, Heck AJ: Next-generation proteomics: towards an integrative view of proteome dynamics. Nat Rev Genet. 2013; 14(1): 35–48. PubMed Abstract \| Publisher Full Text\n\nHindorff LA, Sethupathy P, Junkins HA, et al.: Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc Natl Acad Sci U S A. 2009; 106(23): 9362–9367. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGieger C, Geistlinger L, Altmaier E, et al.: Genetics meets metabolomics: a genome-wide association study of metabolite profiles in human serum. PLoS Genet. 2008; 4(11): e1000282. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHicks AA, Pramstaller PP, Johansson A, et al.: Genetic determinants of circulating sphingolipid concentrations in European populations. PLoS Genet. 2009; 5(10): e1000672. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDemirkan A, van Duijn CM, Ugocsai P, et al.: Genome-wide association study identifies novel loci associated with circulating phospho- and sphingolipid concentrations. PLoS Genet. 2012; 8(2): e1002490. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFu J, Wolfs MG, Deelen P, et al.: Unraveling the regulatory mechanisms underlying tissue-dependent genetic variation of gene expression. PLoS Genet. 2012; 8(1): e1002431. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCookson W, Liang L, Abecasis G, et al.: Mapping complex disease traits with global gene expression. Nat Rev Genet. 2009; 10(3): 184–194. PubMed Abstract \| Publisher Full Text\n\nWestra HJ, Peters MJ, Esko T, et al.: Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet. 2013; 45(10): 1238–1243. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThanabalasingham G, Huffman JE, Kattla JJ, et al.: Mutations in HNF1A result in marked alterations of plasma glycan profile. Diabetes. 2013; 62(4): 1329–1337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu J, Pressoir G, Briggs WH, et al.: A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat Genet. 2006; 38(2): 203–208. PubMed Abstract \| Publisher Full Text\n\nAstle W, Balding DJ: Population structure and cryptic relatedness in genetic association studies. Statist Sci. 2009; 24(4): 451–471. Publisher Full Text\n\nKang HM, Sul JH, Service SK, et al.: Variance component model to account for sample structure in genome-wide association studies. Nat Genet. 2010; 42(4): 348–354. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang Z, Ersoz E, Lai CQ, et al.: Mixed linear model approach adapted for genome-wide association studies. Nat Genet. 2010; 42(4): 355–360. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAulchenko Y, Struchalin M, van Duijn C: ProbABEL package for genome-wide association analysis of imputed data. BMC Bioinformatics. 2010; 11: 134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLippert C, Listgarten J, Liu Y, et al.: FaST linear mixed models for genome-wide association studies. Nat Methods. 2011; 8(10): 833–835. PubMed Abstract \| Publisher Full Text\n\nZhou X, Stephens M: Genome-wide efficient mixed-model analysis for association studies. Nat Genet. 2012; 44(7): 821–824. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSegura V, Vilhjalmsson BJ, Platt A, et al.: An efficient multi-locus mixed-model approach for genome-wide association studies in structured populations. Nat Genet. 2012; 44(7): 825–830. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSvishcheva GR, Axenovich TI, Belonogova NM, et al.: Rapid variance componentsbased method for whole-genome association analysis. Nat Genet. 2012; 44(10): 1166–1170. PubMed Abstract \| Publisher Full Text\n\nFabregat-Traver D, Bientinesi P: Applicationtailored linear algebra algorithms: A search-based approach. Int J High Perform Comput Appl. 2013; 27(4): 425–438. Publisher Full Text\n\nFabregat-Traver D, Aulchenko Y, Bientinesi P: Solving sequences of generalized least-squares problems on multi-threaded architectures. Appl Math Comput. 2014; 234: 606–617. Publisher Full Text\n\nAstle W, Balding DJ: Population structure and cryptic relatedness in genetic association studies. Ph.D. thesis, The University of London. Statist Sci. 2009; 24(4): 451–471. Publisher Full Text\n\nAulchenko YS, Ripke S, Isaacs A, et al.: GenABEL: an R library for genome-wide association analysis. Bioinformatics. 2007; 23(10): 1294–6. PubMed Abstract \| Publisher Full Text\n\nSuhre K, Shin SY, Petersen AK, et al.: Human metabolic individuality in biomedical and pharmaceutical research. Nature. 2011; 477(7362): 54–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoring HH, Curran JE, Johnson MP, et al.: Discovery of expression QTLs using large-scale transcriptional profiling in human lymphocytes. Nat Genet. 2007; 39(10): 1208–1216. PubMed Abstract \| Publisher Full Text\n\nLonsdale J, Thomas J, Salvatore M, et al.: GTEx Consortium. The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013; 45(6): 580–585. 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\nShabalin AA: Matrix eQTL: ultra fast eQTL analysis via large matrix operations. Bioinformatics. 2012; 28(10): 1353–1358. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nConneely KN, Boehnke M: So many correlated tests, so little time! Rapid adjustment of P values for multiple correlated tests. Am J Hum Genet. 2007; 81(6): 1158–1168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi MX, Gui HS, Kwan JH, et al.: GATES: a rapid and powerful gene-based association test using extended Simes procedure. Am J Hum Genet. 2011; 88(3): 283–293. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan der Sluis S, Posthuma D, Dolan CV: TATES: Efficient multivariate genotype-phenotype analysis for genome-wide association studies. PLoS Genet. 2013; 9(1): e1003235. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFusi N, Stegle O, Lawrence ND: Joint modelling of confounding factors and prominent genetic regulators provides increased accuracy in genetical genomics studies. PLoS Comput Biol. 2012; 8(1): e1002330. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFusi N, Lippert C, Borgwardt K, et al.: Detecting regulatory gene-environment interactions with unmeasured environmental factors. Bioinformatics. 2013; 29(11): 1382–1389. PubMed Abstract \| Publisher Full Text\n\nShen X, Alam M, Fikse F, et al.: A novel generalized ridge regression method for quantitative genetics. Genetics. 2013; 193(4): 1255–1268. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFabregat-TRaver D, Sharapov SZ, Hayward C, et al.: OmicABEL software for genome-wide association studies. Zenodo. 2014. Data Source\n\nChen W, Abecasis G: Family-based association tests for genomewide association scans. Am J Hum Genet. 2007; 81(5): 913–926. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDongarra J, Croz JD, Hammarling S, et al.: A set of level 3 basic linear algebra subprograms. ACM Trans Math Softw. 1990; 16(1): 1–17. Publisher Full Text\n\nAnderson E, Bai Z, Bischof C, et al.: LAPACK Users’ Guide. Philadelphia, PA: Society for Industrial and Applied Mathematics, third edition. 1999. Publisher Full Text\n\nVitart V, Biloglav Z, Hayward C, et al.: 3000 years of solitude: extreme differentiation in the island isolates of Dalmatia, Croatia. Eur J Hum Genet. 2006; 14(4): 478–487. PubMed Abstract \| Publisher Full Text\n\nRudan I, Marusić A, Janković S, et al.: “10001 dalmatians:” Croatia launches its national biobank. Croat Med J. 2009; 50(1): 4–6. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "5981", "date": "29 Aug 2014", "name": "Dirk Jan de Koning", "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 manuscript shows a computational approach that enABELs the use of linear mixed mixed models for large pedigrees, large numbers of markers and large numbers of (omics) traits. While the work is interesting, several things need to be clarified before the work can be of interest to the wider community. I assume that the manuscript is aimed at the potential users: researchers in genetics with an interest to analyze their 'big data' in a correct fashion and in a repeatable fashion. The manuscript is compact but possibly too compact in places. I would like the following points to be addressed: Accuracy of estimates and comparison of results. The authors make a big deal about comparing the approaches in terms of computing time. They should also be compared systematically in terms of type I error, accuracy of estimates etc. The computational approximations that are made must come at some cost and as researchers we need to know this cost. There are some token comparisons in the supplementary tables but these deal with heritability and P values. Not with the actual SNP effects which are what we want from these analyses. Some scatter plots between different methods may be helpful. We need more detail about the simulations. You simulate different population sizes but you must clarify the family structures and the family complexity. How many generations? All family sizes equal? etc. Also how you simulate marker data in terms of historical population size, lD structure etc. Likewise for correlation structure among the multiple traits. In many places, variables are used before they are defined. p appears in Table 1 but is only clarified later in the text. The tables are currently not readable on their own and the acronyms and variables should be explained within the tables. In general, the mathematical notation as well as the language in general needs a bit of work. Some sentences are too long with too many clauses, for example CLAK is not defined at first use. X and y are linearly transformed: should you not introduce new variables for the transformed data?Some parts of result should be in the Materials and Methods.Figure 3C could have a logarithmic Y-axis.Figure 3D could omit line for CLAK-Eig and CLAK-Chol is unlabeled.The abbreviation in Table S3 cer-cer lpc, pls spm pc pe etc. makes no sense to me. They are different trait classes but still the table adds very little. The manuscript has the promise to introduce an interesting new computational approach, but we need more details to make up our mind.", "responses": [] }, { "id": "5916", "date": "05 Sep 2014", "name": "Peter Holmans", "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 introduces a new software tool (omicABEL) for performing rapid association analyses between SNP data and high-dimensional \"omics\" data (e.g. gene expression) using mixed models. This is an extremely important development for the following reasons:Identification of genetic variants associated with gene expression (eQTLs), methylation (mQTLs) etc is becoming ever more important in understanding the biology of complex genetic disorders. It is desirable, when performing association analyses, to allow for population stratification and/or cryptic relatedness among members of the sample. Mixed models are an effective way of doing this. The high dimensionality of the \"omics\" data (tens or even hundreds of thousands of measurements) has made the application of mixed model association approaches computationally prohibitive up to now. In addition to detailing an application of considerable practical utility, the article is also clearly written and mathematically rigorous.", "responses": [] }, { "id": "5915", "date": "11 Feb 2015", "name": "Bertram Muller-Myhsok", "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 very well written report on a very important topic of very high impact in the actual of omics and multi-level omics data, allowing beyond sheer analysis (which in large data sets is very demanding indeed in terms of runtime and memory) many more questions answered also for methods research.The methods put forward present a very major step forward for this type of analysis. The methodology is very sound and promising. A tool to actually use is provided, which is also not always the case.The authors are to be commended for this work.", "responses": [] } ]	1	https://f1000research.com/articles/3-200
https://f1000research.com/articles/3-184/v1	05 Aug 14	{ "type": "Research Article", "title": "Actinic keratosis on the rise in Australia", "authors": [ "Eshini Perera", "Sean McGuigan", "Rodney Sinclair", "Sean McGuigan", "Rodney Sinclair" ], "abstract": "Objectives: To report the burden and cost of actinic keratosis (AK) treatment in Australia and to forecast the number of AK treatments and the associated costs to 2020.Design and setting: A retrospective study of data obtained from medicare Australia for AK treated by cryotherapy between 1 January 1994 and 31 December 2012, by year and by state or territory.Results: The total number of AK cryotherapy treatments increased from 247,515 in 1994 to 643,622 in 2012, and we estimate that the number of treatments will increase to 831,952 (95% CI 676,919 to 986,987) by 202. The total Medicare Benefits Schedule (MBS) benefits paid out for AK in 2012 was $19.6 million and we forecast that this will increase to $24.7 million by 2020 (without inflation).Conclusion: The number of AK cryotherapy treatments increased by 160% between 1994 and 2012. we forecast that the number of treatments will increase by 30% between 2012 and 2020. The rates of non-melanoma skin cancer (NMSC) and AK appear to be increasing at the same rate. During the period 2010 to 2015 AK is anticipated to increase by 17.8% which follows a similar trend to published data that forecasts an increase in NMSC treatments of 22.3%.", "keywords": [ "Actinic keratosis (AK)", "is one of the most common1 and most expensive skin diseases to treat. This places a high burden on the Australian population", "health care system and government. Up to 60% of Australians in subtropical Queensland over the age of 40 have AK2. AK lesions are rough scaly and generally less than 1 cm in diameter. They are caused by chronic sun exposure and prevented by regular sunscreen use3. Multiple lesions are common. Some lesions regress spontaneously", "but others progress to invasive squamous cell carcinoma (SCC). Hypertrophic AK may contain occult foci of SCC4. A systematic review of the natural history of AK estimated the risk of malignant transformation to range from 0% to 0.0075%/lesion/year5. In individuals with a history of non-melanoma skin cancer (NMSC)", "the risk is up 0.53% per lesion per year6. Organ transplant recipients on systemic immuno-suppressive therapies are at increased risk of malignant transformation7." ], "content": "Introduction\n\nActinic keratosis (AK), is one of the most common1 and most expensive skin diseases to treat. This places a high burden on the Australian population, health care system and government. Up to 60% of Australians in subtropical Queensland over the age of 40 have AK2. AK lesions are rough scaly and generally less than 1 cm in diameter. They are caused by chronic sun exposure and prevented by regular sunscreen use3. Multiple lesions are common. Some lesions regress spontaneously, but others progress to invasive squamous cell carcinoma (SCC). Hypertrophic AK may contain occult foci of SCC4. A systematic review of the natural history of AK estimated the risk of malignant transformation to range from 0% to 0.0075%/lesion/year5. In individuals with a history of non-melanoma skin cancer (NMSC), the risk is up 0.53% per lesion per year6. Organ transplant recipients on systemic immuno-suppressive therapies are at increased risk of malignant transformation7.\n\nIndications for treatment include prophylaxis of SCC, uncertainty of diagnosis, symptomatic relief and to improve cosmetic appearance. Detectable lesions are frequently associated with alterations of the surrounding skin. These surrounding subclinical lesions, or field changes, can be identified histologically and also clinically with the application of topical field therapy for AK8.\n\nCryosurgery with liquid nitrogen is the most widely used therapy. Clinical response varies with the duration of application, as does morbidity. One study reported cure rates of 39% for freeze times less than 5 seconds, 69% for freeze times between 5 and 25 seconds and 83% for freeze times greater than 25 seconds9.\n\nMedicare Australia is a universal health insurance scheme that reimburses patients for medical fees incurred. Benefits are paid according to the Medicare Benefits Schedule (MBS) with respect to each individual Medicare item number. The item number for treatment of 10 or more solar keratosis by cryosurgery is 30192. There is no item number for treatment of fewer than 10 AK and therefore no reimbursement for cryotherapy treatment of less than 10 AK.\n\nField therapy is most useful for multiple AK lesions. Agents approved by the Therapeutics Goods Administration for field therapy include topical 5-Fluorouracil (5-FU), diclofenac sodium, inguenol mebutate and photodynamic therapy (PDT). Field therapy is not subsidized by the PBS and does not contribute directly to the cost to Government of treatment of AK in Australia.\n\nThere is no current published estimate of the direct cost of AK to the Australian government. We aimed to use Medicare data to report the annual numbers of AK treatments between 1994 and 2012, calculate the direct costs associated with these treatments and predict the numbers and costs from 2013 to 2020.\n\nThe risk of AK increases with age and the ageing of the Australian population may increase the burden of AK on the Australian health system.\n\n\nMethods\n\nMedicare Australia issues the total number of claims and benefits paid for each Medicare item number according to the Medicare Benefits Schedule (MBS)10. The total number of claims and cost of Medicare benefits for the treatment of 10 or more premalignant lesions with an ablative therapy (MBS item number 30192) was obtained for the period 1994 to 2012 inclusive10.\n\nThe historical Medicare benefit payment data were inflated to 2012 Australian dollars, using the Reserve Bank of Australia Inflation calculator11. For each of the SK numbers and costs historical data sets, three Auto-regressive Integrated Moving Average (ARIMA) models12 were fitted and forecasts produced for the years 2013 to 2020. The models were ARIMA(1,1,1), ARIMA(0,1,1) and ARIMA(1,1,0). For each forecast series, the best fitting model was chosen on the basis of the Akaike Information Criterion (AIC)13. For both time series the AIC differences between models were extremely small and the model forecasts very similar.\n\nTotal numbers of non-melanoma skin cancer (NMSC) treatment was obtained from a study conducted by Fransen et al.14. These are illustrated in the graph in Figure 1 to compare the trends of AK to NMSC treatment. All statistical analysis was carried out with JMP® v 9.0.0.\n\nEstimated treatment numbers from the period of 1994–2020. Estimated treatment numbers with 95% CI from 2013–2020 is shaded in grey.\n\n\nResults\n\nThe total number of item 30192 treatment services billed to Medicare in 1994 was 247,515, increasing to 643,622 in 2012. This represents an increase of 160%. Rates remained relatively stable during the period of 2005–2009 and then began increasing again after 2010. It was estimated that the cost of treatments would reach $24.7 million in 2020 for roughly 831,953 treatment services – an increase of 29% over the next 8 years (Table 1 and Figure 1 and Figure 2).\n\nEstimated costs with 95% CI from 2013–2020 is shaded in grey.\n\nThe total number of AK treatments billed to Medicare showed a similar increase over time to the increase seen in total NMSC treatments over the same period of time14 (Figure 3). NMSC treatment services continued to rise during 2005–2009 while AK treatment services remained stable.\n\nTotal number of non-melanoma skin cancer treatment for 1997 to 2012 shown in dotted lines.\n\nThe total number of treatment services billed for each of the Australian states and territories is illustrated in Figure 4. Approximately 293,933 treatment services were billed in NSW in 2012 and 184,233 services in Queensland compared to 1,548.\n\nThe number of AK treatments per 1,000 head was the highest in New South Wales compared to Queensland (Table 2). The number of AK treatments per 1,000 people in the Northern Territory was remarkably low, almost comparable to the number of AK treatments seen in Tasmania.\n\nThe total Medicare benefits paid for the 30192 item number in 1994 was $5.6 million and $19.6 million in 2012. This represents an increase of 250%. NSW and Queensland received the highest amount of benefits at $9 milllion and $5.6 million respectively) while NT received the lowest amount of Medicare benefits at $49 thousand (Figure 5).\n\n\nDiscussion\n\nAK is a chronic disease. Patients with multiple AK require multiple treatment episodes. The number of item 30192 treatments for AK has increased by 160% since 1994 and is predicted to increase by a further 29% between 2012 to 2020. NMSC has similarly been increasing over time. During the period 2010 to 2015 AK is anticipated to increase by 17.8% compared with an increase in NMSC treatments of 22.3% which was predicted by Fransen et al.14.\n\nInterestingly, the rates of ablative treatment for premalignant lesions of the skin remained virtually stable during the period 2005 to 2009, whilst the rates for NMSC treatments continued to rise during this period. One possible explanation for the trend in AK billing treatments is the introduction of new field treatment options during this period. The Therapeutic Goods Administration approved PDT with Methyl aminolevulinate (Metvix®) for the treatment of AK in 2003 and Imiquimod (Aldara®) for treatment of AK in 2004.\n\nTopical 5-FU and PDT are funded by the RPBS but not the PBS for the treatment of AK in repatriation members15. The contribution of these agents to the overall cost to Government of treament of premalignant skin conditions is therefore negligable.\n\nAlso notable is the significantly higher number of AK treatment services, provided by NSW compared to Queensland. This is particularly surprising as the number of NMSC treatments recorded in Queensland continued to increase at a faster rate and the number of NMSC treatments was greater than in NSW after 201110.\n\nNorthern Territory (NT) had the lowest number of AK treatments per 1,000 a figure that was comparable to treament numbers seen in Tasmania. It was expected that the number of AK treatments per head would have been similar to that of Queensland because of the association between AK and ultravoilet (UV) light and the higher levels of UV exposure in NT. Aboriginal and Torres Strait Islanders make up 30% of the population in the NT16. Although there is limited literature on the incidence of AK in the Indigenous population it is likely that the incidence of AK would be less that the Caucasian population as the incidence rates of NMSC and melanomas are reduced in the Indigenous population17, however this still does not account for the lower than expected treatment numbers. Most of the NT is considered rural or remote under the Australian government’s remoteness classification18 and the AK treatment numbers may reflect the reduced access to healthcare in the NT. Furthermore, the lower number of AK treatments may also reflect the behavioural patterns of health seeking in populations living in remote areas.\n\nThere were three major limitations of the study. The data set did not include data on cryotherapy item numbers charged to the Department of Veterans affairs; the data did not include treatment of less than ten lesions; in addition the 30192 item number may include other premalignant conditions of the skin, however we have assumed that AK are the predominant skin lesion. Some AK may clinically resemble SCC and be treated surgically. We have not calculated the costs of skin surgery of AK. Our calculations did not include any patient co-payments costs incurred or the cost of the consultation.\n\nThe cost to the government and the cost to the community of AK is increasing. We predict that the number of AK treatments provided by the government in 2020 will be 831,953 costing an estimated 24.7 million dollars.\n\nAn estimated 60% of SCCs develop from AK and the remaining 40% develop de novo19. Given the high risk of malignant transformation treatment of early treatment and prevention of AK is prudent.\n\nFuture research should focus on identifying future AK trends, cost effectiveness of field treatments versus cryotherapy for AK, the association between AK and SCCs and the role of field treatments for clinical and subclinical AKs in preventing SCCs.\n\n\nData availability\n\nThe data in this study and the study by Fransen et al. can be accessed via Medicare Australia Statistics. http://www.medicareaustralia.gov.au/statistics/mbs_item.shtml. The item number used was 30192; the dates used were 1994–2012.", "appendix": "Author contributions\n\n\n\nThe authors Eshini Perera and Rodney Sinclair were involved in writing and editing the draft of this manuscript. Sean McGuigan was responsible for the methodology and statistical calculations used in this manuscript. All authors of this research paper have directly participated in the planning, execution, or analysis of this study. All authors of this paper have read and approved the final version submitted.\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\nNeidecker MV, Davis-Ajami ML, Balkrishnan R, et al.: Pharmacoeconomic considerations in treating actinic keratosis. Pharmacoeconomics. 2009; 27(6): 451–64. PubMed Abstract \| Publisher Full Text\n\nFrost CA, Green AC, Williams GM: The prevalence and determinants of solar keratoses at a subtropical latitude (Queensland, Australia). Br J Dermatol. 1998; 139(6): 1033–39. PubMed Abstract \| Publisher Full Text\n\nThompson SC, Jolley D, Marks R: Reduction of solar keratoses by regular sunscreen use. N Engl J Med. 1993; 329(16): 1147–51. PubMed Abstract \| Publisher Full Text\n\nScarfe C, Sinclair R: Actinic Keratosis. In: Lebwohl M, Heymann W, Berth-Jones J, Coulson I, Editors. Treatment of Dermatological Disease. London: Harcourt Health Sciences. 2002.\n\nWerner RN, Sammain A, Erdmann R, et al.: The Natural History of Actinic Keratosis: A Systematic Review. Br J Dermatol. 2013; 169(3): 502–518. PubMed Abstract \| Publisher Full Text\n\nGupta AK, Paquet M, Villanueva E, et al.: Interventions for actinic keratoses. Cochrane Database Syst Rev. 2012; 12: CD004415. PubMed Abstract \| Publisher Full Text\n\nLindelof B, Sigurgeirsson B, Gabel H, et al.: Incidence of skin cancer in 5356 patients following organ transplantation. Br J Dermatol. 2000; 143(3): 513–9. PubMed Abstract \| Publisher Full Text\n\nBerman B, Bienstock L, Kuritzky L, et al.: Actinic keratoses: sequelae and treatments. Recommendations from a consensus panel. J Fam Pract. 2006; 55(5): S1–8. PubMed Abstract\n\nThai KE, Fergin P, Freeman M, et al.: A prospective study of the use of cryosurgery for the treatment of actinic keratoses. Int J Dermatol. 2004; 43(9): 687–92. PubMed Abstract \| Publisher Full Text\n\nMedicare Australia. Medicare Benefits Schedule item statistics reports. 2013. Reference Source\n\nReserve Bank of Australia. Inflation calculator. 2013. Reference Source\n\nBeran J: Maximum Likelihood Estimation of the Differencing Parameter for Invertible Short and Long Memory Autoregressive Integrated Moving Average Models. J R Statist Soc B. 1995; 57(4): 659–672. Reference Source\n\nHurvich C, Tsai C: Regression and time series model selection in small samples. Biometrika. 1989; 76(2): 297–307. Publisher Full Text\n\nFransen M, Karahalios A, Sharma N, et al.: Non-melanoma skin cancer in Australia. Med J Aust. 2012; 197(10): 565–68. PubMed Abstract \| Publisher Full Text\n\nMedicare Australia. Pharmaceutical Benefits Schedule Item Reports. 2013. Reference Source\n\nAustralian Bureau of Statistics. Regional statistics, Northern Territory. 2013. Reference Source\n\nAustralian Institute of Health and Welfare. Cancer in Australia: an overview, 2008. Canberra: AIHW, 2008. (AIHW Cat. No. CAN 42; Cancer series No. 46). 2013. Reference Source\n\nAustralian Institute of Health and Welfare. Rural, regional and remote health: a guide to remoteness classifications. Canberra: AIHW, 2004. (AIHW Cat. No. PHE 53; Rural Health Series No. 4). 2013. Reference Source\n\nBerman B, Bienstock L, Kuritzky L, et al.: Actinic keratoses: sequelae and treatments. Recommendations from a consensus panel. J Fam Pract. 2006; 55(5): S1–8. PubMed Abstract" }	[ { "id": "5703", "date": "06 Aug 2014", "name": "Luigi 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\nThis is an interesting study dealing with the number of treatments reimbursed by Medicare for AK in Australia over time, and paralleling those treatments with number of treatments for SCC. As discussed by the Authors, the data presented have several limitations mainly connected with the nature of the claims data available (there is a lack of information on treatments of less than 10 AKs, and a lack of information on treatment modalities other than cryotherapy). In addition, a few points are worth of consideration:The Title does not accurately reflect the contents of the paper. I would change it to: \"Costs for the treatment of AKs on the rise in Australia\" The study is restricted to treatment of 10 or more AKs by cryotherapy. Is there any estimate of the number of treatments for modalities other than cryotherapy (including field therapy)? Is there any indication that some lesions were not treated at all? Were repeated treatments of the same lesions counted separately or did they simply sum up to the total? Introduction, 5th paragraph: what does \"PBS\" stand for? Methods, 3rd paragraph: as far as I understand the meaning, the reference should be to Figure 3 and not to Figure 1. Discussion, penultimate paragraph: \"treatment of early treatment and prevention of AK is prudent\" should read \"treatment of early lesions and prevention of AK is prudent\"", "responses": [ { "c_id": "945", "date": "16 Aug 2014", "name": "Eshini Perera", "role": "Author Response", "response": "Thank you for reviewing our article. We have made the following changes to our article, as per your recommendation:The title was amended to reflect the content of the paper.There are no estimates for the number of treatments for modalities other than cryotherapy. This was mentioned in the article. It is possible that some lesions were not treated and this was mentioned in the article. It is possible that some lesions could have been treated twice and this was mentioned in the article. The 3rd paragraph in the methods section was amended so that the meaning reflected what was seen in Figure 1. The error in the discussion paragraph was corrected. We thank you for your suggestions." } ] }, { "id": "5704", "date": "12 Aug 2014", "name": "Giuseppe Argenziano", "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 an interesting study outlining the rising number of treatments of actinic keratoses (AK) and non melanoma skin cancers in Australia. The paper reads well and deals with an important issue.It would be useful to understand whether the increasing number of treated AKs actually corresponds to a decrease of invasive squamous cell carcinomas. If this is not the case, then the burden and cost of AK treatment may be not completely justified. This issue should be shortly addressed in the discussion.Please check in the abstract, results section, \"...by 202\" should read \"2020\".", "responses": [ { "c_id": "944", "date": "16 Aug 2014", "name": "Eshini Perera", "role": "Author Response", "response": "Thank you kindly for reviewing our article. We have amended the article to include your suggestions: Incidence of SCC was briefly mentioned in the discussion. Unfortunately there is no recent data on the incidence of cutaneous SCC. There is also no data which was collected during the time frame of our study. The increasing number of AK treatments could not be correlated to the number of SCC. The abstract was corrected to read 2020." } ] } ]	1	https://f1000research.com/articles/3-184
https://f1000research.com/articles/3-199/v1	20 Aug 14	{ "type": "Software Tool Article", "title": "EpiCollect+: linking smartphones to web applications for complex data collection projects", "authors": [ "David M. Aanensen", "Derek M. Huntley", "Mirko Menegazzo", "Chris I. Powell", "Brian G. Spratt", "Derek M. Huntley", "Mirko Menegazzo", "Chris I. Powell", "Brian G. Spratt" ], "abstract": "Previously, we have described the development of the generic mobile phone data gathering tool, EpiCollect, and an associated web application, providing two-way communication between multiple data gatherers and a project database. This software only allows data collection on the phone using a single questionnaire form that is tailored to the needs of the user (including a single GPS point and photo per entry), whereas many applications require a more complex structure, allowing users to link a series of forms in a linear or branching hierarchy, along with the addition of any number of media types accessible from smartphones and/or tablet devices (e.g., GPS, photos, videos, sound clips and barcode scanning). A much enhanced version of EpiCollect has been developed (EpiCollect+). The individual data collection forms in EpiCollect+ provide more design complexity than the single form used in EpiCollect, and the software allows the generation of complex data collection projects through the ability to link many forms together in a linear (or branching) hierarchy. Furthermore, EpiCollect+ allows the collection of multiple media types as well as standard text fields, increased data validation and form logic. The entire process of setting up a complex mobile phone data collection project to the specification of a user (project and form definitions) can be undertaken at the EpiCollect+ website using a simple ‘drag and drop’ procedure, with visualisation of the data gathered using Google Maps and charts at the project website. EpiCollect+ is suitable for situations where multiple users transmit complex data by mobile phone (or other Android devices) to a single project web database and is already being used for a range of field projects, particularly public health projects in sub-Saharan Africa. However, many uses can be envisaged from education, ecology and epidemiology to citizen science.", "keywords": [ "The increasing availability and decreasing cost of mobile devices running the open-source Android operating system1", "in conjunction with greater geographic penetration of mobile data networks and concurrent increase in bandwidth2", "provides ever-increasing opportunities for the use of such devices in many kinds of data gathering projects. Existing initiatives within this area broadly fall into two groups. Firstly", "those that focus on standard mobile devices and take advantage of standard mobile networks for the transmission of data (e.g.", "using SMS to pass text messages", "as in RapidSMS3 or richer form-based data in EpiSurveyor (now Magpi)4) and", "secondly", "those that take advantage of mobile data networks to enable the passing of data via the web (e.g.", "ODK5 and EpiCollect6)." ], "content": "Introduction\n\nThe increasing availability and decreasing cost of mobile devices running the open-source Android operating system1, in conjunction with greater geographic penetration of mobile data networks and concurrent increase in bandwidth2, provides ever-increasing opportunities for the use of such devices in many kinds of data gathering projects. Existing initiatives within this area broadly fall into two groups. Firstly, those that focus on standard mobile devices and take advantage of standard mobile networks for the transmission of data (e.g., using SMS to pass text messages, as in RapidSMS3 or richer form-based data in EpiSurveyor (now Magpi)4) and, secondly, those that take advantage of mobile data networks to enable the passing of data via the web (e.g., ODK5 and EpiCollect6).\n\nCommon to all initiatives is the definition of the data to be gathered as a ‘form’ which details the questions presented to a user, with the method of form definition varying between initiatives. SMS-based initiatives generally utilise coded responses to compress the answers to questions into short strings, which can be transmitted via SMS. At the other end of the spectrum, and more generally with respect to the initiatives that use mobile data networks for data transfer, more complex single forms can be defined using standard vocabularies (usually eXtensible Markup Language, XML), which are loaded onto a phone, with data gathered and sent centrally for further analysis via the web. The complexity of setup, form definition and functionality available to an end user varies between initiatives; however, common to all is the definition of only a single form per project.\n\nEpiCollect was developed to provide a simple and intuitive method for complete online project creation, data storage and visualisation of aggregated data captured using smartphones based on a single form. Since release, EpiCollect has been used for a broad range of data collection projects using Android smartphones, and the data collection form for a specific project and an associated web application can be generated online by a simple process at the EpiCollect website (www.epicollect.net). Despite its continuing utility, and ease of setup, many data collection projects require considerably more complexity than a single form. For example, to use mobile phones to collect information about the vaccination status of children in schools in a region of a developing country, we would need a form for each school in the area and, linked to each school, a form for each class in the school and, linked to each class, a form for each child in the class. This hierarchy of forms allows data to be collected in a ‘one-to-many’ fashion as required by the project. Additionally, we may need branches in this linear hierarchy of forms so that, if a question on a form is answered ‘yes’, an additional form that needs to be completed one or more times appears on the phone as a branch link.\n\nA greatly enhanced version of EpiCollect (EpiCollect+) has therefore been developed that provides the ability to produce individual data collection forms that are more complex than the single form used in the original version (which is now referred to as EpiCollect or EpiCollect v1), and to link multiple forms together in a linear or branching hierarchy. These new features of EpiCollect+ make it a very versatile tool for data collection via mobile phones. The EpiCollect (and EpiCollect+) app has been developed for any device running the Android operating system (e.g. mobile phone or tablet); however, for convenience we will henceforth refer to ‘mobile phone(s)’.\n\n\nEpiCollect v1 – a framework for simple mobile data gathering projects\n\nWe have previously detailed the overall framework for mobile data gathering using EpiCollect6 and here we briefly detail the online process for project definition and data visualisation that is now available at www.epicollect.net (Figure 1).\n\nWebsite creation is undertaken at www.epicollect.net (A), followed by online form design (B). A project is loaded onto one or more phones and, following data gathering and synchronisation (C), data are visualised using Google Maps and charts at the project website (D). Data can also be entered or downloaded directly via the project website.\n\nEpiCollect v1 projects are hosted using Google’s cloud computing platform AppEngine (GAE)7 primarily for two reasons. Firstly, GAE provides automatic scalability and, secondly, we can readily use Google Accounts authentication (passwords) for assigning ownership to a particular project. Although GAE is used extensively for data storage and application provision, EpiCollect v1 software and documentation are provided to allow project setup on other servers with data sent/retrieved from the mobile app to/from a database set up outside of GAE.\n\nAll methods detailed here are generic and applicable to many kinds of projects. However, for demonstrating how a simple EpiCollect v1 project is setup online using GAE, we describe a hypothetical single form data gathering project where we wish to undertake a survey of schools. Initially, we wish to know five things – the name of each school, the type of school (state, independent etc.), whether the school is single-sex or mixed, and the number of teachers and pupils. Furthermore, we wish to capture an image of each school and record its geographic location.\n\nA) Create. Naming a project and creating a permanent project homepage.\n\nThe ‘Create project’ link at www.epicollect.net generates a prompt for entering a new project name – in this case ‘schoolsdemo’. Following input (and a check to make sure the name is unique), a permanent URL (the project homepage) is generated that includes the project name (as follows):\n\nhttp://www.epicollect.net/project.html?name=schoolsdemo\n\nB) Design. Specifying the survey questions using a ‘drag and drop’ form builder.\n\nA data collection form detailing the questions to be asked of each individual participating in the survey is now generated. The project owner is first prompted to login to epicollect.net using a Gmail account and, henceforth, only this Gmail account holder can amend the project form and edit the project homepage. Once logged in, the project owner is presented with a ‘drag and drop’ form builder, which provides a simple and intuitive way to define the questions presented to a user and consists of three sections (Figure 2). A list of the types of available Form Elements are shown (Figure 2A), and these are ‘dragged’ to a blank form in order to build up the required questionnaire (Figure 2B). Four types of text field are available – ‘Text input’ where a user enters free text, ‘Long text’ which allows a larger number of text characters than the ‘Text input’ option, ‘Select One’ which is equivalent to a dropdown list where choices are defined and a user can select ONLY one answer, and ‘Select Multiple’ which is equivalent to a checkbox list where choices are pre-defined and a user can select one or more answer(s).\n\nForm elements available are listed on the left (A); any number can be dragged to the central section to build up a form (B). For each form element chosen, details are entered on the right (C) including the text presented to a user and possible choices should a pre-defined list of answers be required.\n\nWhen dragging a particular form element onto the form, a user must define the element details that specify the particulars about an instance of that element (Figure 2C). These include the ‘Label’ (which specifies the text displayed to a user answering the questionnaire) and the ‘ID’ (a unique name given to a particular data field and used as a reference when storing the data - the equivalent of a column heading in a spreadsheet or database table). For multiple choice questions (Select One and Select Multiple), the available choices can be defined at this point where the ‘Name’ specifies what is shown to the user and the ‘Value’ indicates what is stored in the database when the particular option is chosen by a user. Validation for a particular field can be set as ‘required’ or ‘numerical’ or both, to denote that the question MUST be answered and/or must only contain an integer. Finally, for ‘Select One’ (dropdown) and ‘Select Multiple’ (checkbox) elements, charting can be enabled, so that a pie or bar chart showing the distribution of the answers of users will be shown when the data submitted by mobile phone(s) are viewed at the project website. There is no limit on the number and type of questions that can be specified in a form, although the number of chartable fields is currently limited to three. In this way, forms containing many questions can be built up easily on-line and tailored to the data gathering needs of the project owner. On completion of a form, the project is ready to be loaded onto one or more mobile phones.\n\nC) Collect. Loading a project onto a mobile phone(s) and collecting data.\n\nThe EpiCollect v1 mobile app is first installed on each Android phone to be used in the survey (via Google Play or direct download from the EpiCollect website). From the mobile app’s homescreen, entering the project name in the ‘Load Project’ menu option (e.g., ‘schoolsdemo’) requests the project definition file from GAE. Once loaded, three main options are now available from the homescreen - ‘New entry’, ‘List entries’ and ‘Display map’. ‘New entry’ presents the data collection screen. Each new entry is assigned a unique ID and automatically date and time stamped. A location (if GPS or wireless is enabled on the phone), and a photo (captured using the phone’s camera), can be attached to an entry and then the questions defined in the survey form are answered. Each completed entry is stored in the phone’s database, allowing many entries to be collected on one handset (only limited by memory). Synchronising data to the project website is achieved by listing the entries and tapping ‘synchronise’, which sends a copy of the survey data gathered on the mobile phone to GAE, which can then be viewed at the project website along with entries already submitted from other phones being used in the project. Multiple projects can be loaded on a single phone and the project to be undertaken can be switched from a dropdown list on the mobile app’s homescreen.\n\nData connectivity (e.g., 2/3/4G or wireless) is only required when initially loading the project into the app and when synchronising the data with the project database. Importantly, it is NOT required for entering data (GPS does not require data networks), allowing data gathering in remote areas with later synchronisation when data connectivity is available. Furthermore, data from GAE can be requested to the mobile app so that data previously collected by the user, and/or other data collectors, can be viewed on any handset being used in the project.\n\nD) Visualise. Viewing data at the project homepage.\n\nThe data collected from one or more phones during the survey can be viewed via the project homepage in two ways. Firstly, data can be listed in a tabular view, with columns based on the defined form fields, and all data can be downloaded in a number of formats (CSV/XML) for use and further interpretation outside of EpiCollect. Secondly, data can be viewed using a generic mapping interface (Figure 3). Briefly, all entries where location data have been captured (latitude/longitude via either GPS or wireless) are listed (Figure 3A) and can be viewed on a Google Map (or within Google Earth). The Google Map interface (Figure 3B) includes charts (pie and/or bar) for displaying data from fields defined as ‘chartable’ when building the project form (Figure 3C), and a temporal ‘slider’ allows only those points gathered within a particular time period to be viewed (Figure 3D). Data that are chartable can also be filtered based on choosing a particular ‘answer’ to a question that was presented to the users, so that only those entries where this answer is given are shown on the map (Figure 3E).\n\nData shown are from a demo-project, which comes preloaded with the EpiCollect v1 mobile application. All data points for a project are listed at the top left (A) and their position (if recorded) is plotted on a Google Map (B). Any fields that were specified to be charted when designing a form are displayed in the Chart View (C). A temporal slider allows the data collected within a specified time period to be displayed (D). Data can be filtered based on any ‘Select one’ or ‘Select multiple’ field, allowing the display of only those entries with the specified value(s) (E).\n\n\nUsing XML to define projects\n\nWithin EpiCollect we have focussed on complete generality allowing the design of any kind of project that would benefit from the collection of questionnaires along with location and photographic data. This generality is achieved by defining a project description using XML.\n\nFor EpiCollect v1, we defined a set of XML tags (a vocabulary;8) that allow all elements of project description to be defined in a single document. Within the online interface, the XML details are hidden from the user. However, in brief, when defining a form, the ‘drag and drop’ form builder produces an XML description of the form, including the name of the project. This XML file is then utilised both on the GAE server (for production of the generic map interface/tabular view when data is viewed) and within the mobile app when a project is loaded, to indicate the questions to be presented to a user and the location of the GAE server (tagged by project name) to send/retrieve data to/from.\n\nWe utilise XML for a number of reasons. Firstly, it allows a standard and flexible vocabulary for defining projects; secondly, it allows users to host and undertake their own data gathering projects using their own databases and servers. By default, when loading a project on the EpiCollect mobile app, a request is sent to GAE for retrieval of the corresponding project description (in XML). However, XML can be loaded from any web location by typing in a fully resolved URL. Thus, by defining a project in XML (including server location and scripts for handling data exchange), and hosting on a local server, users can utilise the mobile app for data gathering and retrieve and store all data on their own servers. Further instructions are available in the developer’s section at epicollect.net.\n\nBecause XML is extensible, we can expand our vocabulary as and when required. As long as the software that utilises the XML ‘understands’ the vocabulary, tags can be defined which address greater complexity. Expansion of this principle (additional tags, interpreted by server and mobile software) forms the basis of progressing from EpiCollect v1 to EpiCollect+.\n\n\nFrom EpiCollect to EpiCollect+\n\nWhile EpiCollect v1 may prove adequate for many types of projects, in other cases, a greater control over the flow of questions becomes necessary and projects may require the collection of multiple linked forms within a single project. Similarly, the limit (and requirement) for the collection of a single image and a single GPS may be restrictive and does not take advantage of other media types available to phones running the Android operating system. For example, the camera can be used to capture short videos, or for the extraction of data encoded in barcodes, the microphone can be used to capture sound clips. Moreover, it is possible, through Bluetooth bridges, to connect to other ‘sensors’ for the acquisition of data additional to those provided from mobile phones (e.g., Arduino9).\n\nWe have therefore developed EpiCollect+, which uses a greatly enhanced XML vocabulary for project description than that used for EpiCollect v1, allowing multiple text and/or media fields, multiple linked forms, and increased validation and control over form logic, to be specified easily. Furthermore, as described for EpiCollect v1, we include a ‘drag and drop’ interface for EpiCollect+ project definition, providing a complete online solution to complex project definition and data gathering.\n\nIn brief, the EpiCollect+ XML description allows the definition of projects that contain one or more linked data forms (linked either hierarchically, in a one-to-many fashion, or branched) and each form can contain any number of the standard ‘text’ and ‘media’ fields (Figure 4). We describe the enhanced features of EpiCollect+ by first outlining the increased complexity available for the individual data collection forms, and then the hierarchical and branch linking of multiple forms.\n\nText fields mirror those available in EpiCollect v1 and media fields include video clips, sound clips and barcode scanning. Any number (0, 1 or many) of any of these fields can be included in a form.\n\n\nText fields\n\nAs well as the text fields described for EpiCollect v1 (‘Select one’ fields are henceforth referred to as ‘Dropdowns’ and ‘Select multiple’ as ‘Checkboxes’), EpiCollect+ allows the use of radio buttons which are conceptually similar to a dropdown list, in that choices are predefined and a user is allowed only one choice; however all options are displayed on the phone’s screen at one time. Within an individual form, any number of such fields can be specified and are presented to a user in the order defined. However, defining form logic (e.g., setting skip patterns) and validation of the data entered, allow greater control over the data entered (see below – Form data logic and validation).\n\n\nMedia fields\n\nXML tags have been specified that allow the capture of a number of additional media types to those available in EpiCollect v1:\n\nLocation: The phone’s position obtained using built-in GPS or via network or wireless positioning. The latitude, longitude, altitude and accuracy (in metres) are recorded.\n\nImage: An image from the phone’s picture library or an image captured using the phone’s camera.\n\nVideo: A video clip, captured using the phone’s camera.\n\nAudio: A sound clip, captured using the phone’s microphone.\n\nBarcode: Extraction of data encoded within a 2D barcode, using the phone’s camera, is entered into a text field within the form (utilising either Google’s built-in barcode scanner within recent versions of Android or the ZXing library10, which users are prompted to install if no barcode reader is detected).\n\nAny number of each media type (0, 1 or more) can be defined within a form and all data are synchronised to the server and available for download/viewing.\n\n\nData logic and validation\n\nEpiCollect+ provides control over the flow of questions within a form using the following methods:\n\nForm logic: This allows the definition of questions to ‘jump to’, should a particular answer be given to a particular question. For example, if the answer to Q1 is A, jump to Q5; for all other choices continue to Q2. Multiple clauses within a jump can be defined to allow greater flexibility. For example, if the answer to Q1 is A or C, jump to Q5, or if the answer is B, jump to Q7, else continue to Q2. A common example where jumps are used is to record a value for a question with defined choices (i.e. dropdowns, radio buttons) where one of the choices is ‘other’. Thus, if Q1 is ‘What is your favourite colour?’, with the options ‘blue’, ‘red’, ‘green’ and ‘other’, a jump to Q3 could be used if one of the pre-defined colours was selected. If ‘other’ was chosen the user would proceed to Q2 (containing a simple text box), where they would be asked to enter their favourite colour (the ‘other’ colour that was not in the initial list) before proceeding to Q3.\n\nData validation: Data validation within the description of a form allows the setting of rules that attempt to minimise the opportunity for data entry errors or entry of incorrect data types. Free text is open to the potential for spurious data input and a number of methods are provided for specifying data validation. These include requiring fields and adherence to a particular data type; e.g., integer only, dates and times with one of a number of specified formats, such as dd/mm/yyyy or mm:ss, and minimum and/or maximum accepted values for numeric fields allowing the definition of ranges. Furthermore, date and time fields can take advantage of graphical ‘widgets’ to provide scrollwheels for entering dates.\n\nRegular expression pattern-matching is also fully implemented within EpiCollect+, allowing the definition of a particular string pattern for matching against a user’s input, with warnings should the data entered conflict with the pattern specified. Such validation is suited to free text but can also be applied to other field types. For example, regular expressions can be used in conjunction with barcodes. Should a barcode be known to encode data starting with a particular series of characters, this can be specified using a regular expression and a warning returned should the wrong type of barcode be scanned. Any text field (and also barcode fields) can also be flagged for ‘double entry’ and when users are presented with the question (or request to scan the barcode), they are prompted to enter the data twice. Cross-checking ensures that both entries are identical before proceeding to the next question (as commonly used within web forms for passwords that have to be entered twice and are cross-checked).\n\n\nLinking forms\n\nIn EpiCollect+ multiple single forms can be linked together within a project in two ways – hierarchical and branching.\n\nHierarchical linking: The schools project mentioned above is used to demonstrate the general principles of form linking available in EpiCollect+. Within this schools project, we intend to survey schools from across a defined area and to record details about each school. However, in addition, we now wish to collect information about each class (class size, class teacher etc.), and about each pupil in each class. Furthermore, for each pupil, we wish to collect two media types. Firstly, a short audio clip of the pupil pronouncing their own name and, secondly, we wish to take a blood sample, which will be processed by a laboratory to obtain their blood group, with these data subsequently being added to the survey. For the audio, we will use the phone’s microphone to record a short clip and, for the blood sample, we will use pre-printed barcodes and ‘scan’ a barcode, associating the code to the entry before sticking the printed barcode to a blood sample tube.\n\nTo undertake data gathering, we could design a single form that, for each pupil, includes details about their class and their school, but this would lead to vast repetition of data entry, and there is a natural hierarchy within the project that lends itself to three forms being used in a ‘one-to-many’ fashion. Each school has many classes and each class has many pupils. If we define this hierarchy from top to bottom as school->class->pupil, we can design a single form for each of these three levels, and link them together so that the one-to-many hierarchy is preserved. Within EpiCollect+ this is achieved by adding a new form and defining a ‘key’ field that is used to link data in one level of the hierarchy to the level above and/or below (Figure 5A). Each level is defined as a single form (including any number of text and media fields) and linked by key fields, allowing data gathering to occur in the one-to-many fashion (Figure 5B). A key field should contain data that are unique to each instance of a form. For example, within the schools project, we could define ‘school name’ as the key field in the school form (each school will have a unique name). This is linked to the class form so that all class forms contain the name of the school they are associated with. Similarly, we could define ‘class name’ (each class will have a unique name) as the key field in the class form which is included in all pupil forms, linking specific pupils to a class within a school.\n\nForms can be linked hierarchically (A), allowing a ‘one-to-many’ relationship between the data gathered (B). Forms can also be linked as a branch (C) allowing, at any point within a form, ‘branching’ out to fill in one or more entries for a branch form.\n\nBranch linking: A second method of linking forms is provided by the branch form type (Figure 5C). Within the schools project we may wish to record, within each class form, the number of absences taken by a teacher during the previous academic year and, for each absence, to capture details specific to each absence (e.g., reason for absence, number of days etc). Initially, the number of absences per teacher is unknown and may require no additional forms (a teacher has not been absent) or many ‘absence’ forms (a teacher has been absent more than once). This uncertainty in the number of forms required for each teacher precludes inclusion in a hierarchy as this would require at least one entry per teacher and, therefore, does not logically fall into the one-to-many linear hierarchy. To address these kinds of extra data forms, we allow the definition of a ‘branch form’. A question within the initial form (e.g., ‘Please record absences’) causes the branch to be triggered and presented to the user. In this fashion, when a user fills in details of each class there is a question about the number of absences of their teacher. The response to this question triggers the ‘absence’ branch form (if required) and the required number can be filled in before continuing with the next question on the class form. If there are no absences, the class form continues automatically to the next question. Branch forms can be defined at any point within any form within a hierarchy and, furthermore, multiple branches can be defined within a single form.\n\nDefining a project in this fashion means that initially, we simply describe the fields (text and/or media) for each type of form, and link them together to provide the overall hierarchy that is required.\n\n\nDefining keys and synchronising data with the central database\n\nAs in EpiCollect v1, the data collected on each mobile phone are synchronised with the central server. However, in EpiCollect+, data are sent sequentially from the highest level in the hierarchy downwards to preserve the integrity between ‘keys’ in forms at all levels. An entry from a form lower in the hierarchy must have an entry ‘above’ it for the entry to be successfully entered and synchronised.\n\nWhen designing a form, a user can choose to define their ‘key’ field or we offer the ability to automatically generate a key, which can be displayed or hidden from the user on the form. The auto-generated key consists of the phone’s universal unique identifier, concatenated with the exact time the entry is taken in milliseconds. This precludes the possibility of data being entered more than once.\n\n\nUndertaking a project with EpiCollect+\n\nAs in EpiCollect v1, all elements of project design in EpiCollect+ are undertaken online using a ‘drag and drop’ form builder and the project is loaded onto one or more phones. When starting data collection, forms are listed in the order defined in the hierarchy and must be entered in this fashion. In the schools project, the forms are listed in the order ‘school’, ‘class’, ‘pupil’ (See Figure 6). Once the top level ‘school’ form is filled in, a menu is shown which offers three choices. 1) ‘Add another school’, allowing a new entry to be added. 2) ‘List/Sync schools’, which allows all school entries to be viewed and synchronised with the central server. 3) ‘Add class to school’, which allows the user to fill in a form immediately below in the hierarchy, which will be associated with the school through the school forms key field. Each level of the linear hierarchy works in this way – i.e., on completion of a form entry, users are prompted to add entries in the level below in the hierarchy.\n\nLoading the Schools project lists three forms in the ‘Select Form’ menu in the order ‘School’, ‘Class’ and ‘Pupil’ (A). Selecting the top level form (School) presents the School menu (B), containing options for viewing or adding data. ‘Add School’ presents the user with the questions specified (text and/or media) allowing single or multiple school entries to be added (C). After completion of an entry, further schools can be added or a ‘Class’ (representing the second level in the hierarchy) can be added to a particular school (D). Similarly, after completion of a Class entry a user may choose to add another Class or drop to the third level of the hierarchy and add one or more Pupils to a particular Class (E). The presence of a branch ‘Sickness’ form in the Class form can be utilised to add any number of Sickness forms to a particular Class as described in the text (F). Synching can be undertaken at any point within the form hierarchy.\n\nWhen completing a form on a mobile phone, text fields are completed via the keyboard and, for media fields, the user is prompted to record the media specified. For example, for any location field (i.e., GPS), the user is prompted to ‘tap-to-set location’ and if a GPS or wireless signal is available the screen will be updated with location details (latitude/longitude and accuracy). For picture or video fields, the phone’s camera is launched (or the user can select from the phone’s library of images/videos). For audio clips, audio controls allow ‘stop/start’ for recording a short clip. For barcode scans, the phone’s camera is launched, allowing the user to point the camera at a barcode and, once recognised, the data encoded are extracted and placed in a textbox on screen.\n\nSynchronising data can be undertaken either after each entry is completed or all at once from the app’s homescreen. All textual data, from all levels in a project hierarchy, are synchronised in this fashion. For photos, a thumbnail is sent to the server by default. Due to the potentially large amount of data transfer when synchronising multiple photos, video and audio clips, we provide a menu option that allows these data to be synchronised separately. This is provided to give the user the choice of whether to send all large media data via mobile networks or wireless.\n\nAt the project homepage, as with EpiCollect v1, all data can be viewed in two ways. A list of forms, in hierarchical order, is shown allowing selection of the form entries required. A tabular view of data is shown listing all entries, allowing basic functions such as sorting/searching. For pictures, a thumbnail is present and, if clicked, the full size image is seen (if synchronised – see above). For audio and video recordings, a link is provided to download clips for listening/viewing. For any form that has a location field, a further map view is provided which, as in EpiCollect v1, places all entries on a map (clustered at lower zoom levels), which can be clicked to view further details in a pop-up window. Charts can be viewed for any dropdown, checkbox or radio button field and a temporal slider is available to view entries gathered within a particular time period. For any form, all data are available to download as comma- or tab-separated plain text files or XML which, by default, include URLs to any media fields recorded within a form.\n\nData collection is not reliant on a mobile phone and can be entered at any level within a project’s form hierarchy directly via the web. For data entry via the web, a similar interface is provided to that available on the phone and each question is answered in turn. For location fields, a geo-coding facility is provided that allows an address to be entered and its latitude/longitude retrieved and, for other media fields (picture, video and audio), a file can be uploaded from the user’s computer. Currently, for barcode fields, only the decrypted data can be entered via the web.\n\n\nLoading data from a project database onto mobile phones\n\nWithin any project it is possible to pre-load data from the central database onto a mobile phone at any level within a project hierarchy allowing, for example, the continuation of data gathering from mid-point within a project. For example, within the schools project, we may already know the names and details of many or all of the schools we wish to survey. We may then recruit 15 new data gatherers to aid completion of the project. Rather than re-entering the school information for the new users to subsequently add further classes/pupils, the schools project can be loaded onto their phones and, from the menu, the form level within the hierarchy to pre-load can be selected. Choosing ‘school’ would load all the data on all schools from the central server. Similarly, selecting the ‘class’ level within the hierarchy will load all school and class data. All data within and above levels in the hierarchy are loaded to preserve data integrity.\n\n\nServer software\n\nGeneric server software is provided that allows the hosting, storage and viewing of data collected using the EpiCollect+ mobile app. Briefly, the server software requires three things: a web server (e.g., Microsoft IIS or the open source Apache), a relational database (MySQL or similar) and the server-side PHP scripting language. Installation of such software should be trivial for system administrators and bundled software is available that will install and configure all three components in a single click-to-install executable (e.g. XAMPP11). The only other requirement is that the server is web-accessible with either a static IP address or via a URL. This location is then specified within the phone (through a ‘settings’ option) to indicate to the mobile app where to submit and retrieve data to/from.\n\nThe same project building functionality is provided for any installed server (including the ‘drag and drop’ form builder and also the option to directly upload an XML document) and loading onto the mobile app occurs in the same fashion as a project setup at epicollect.net. When a user selects ‘load project’, and types the name of the project, the mobile app requests the project definition from the server specified. Multiple projects can be setup on a single server and each will have its own URL e.g., www.myserver.com/epicollectplus/projectname where ‘www.myserver.com’ and ‘projectname’ indicate the server location and name of the project. Within the EpiCollect+ server software, user administration allows users to be assigned as managers/curators or users. Managers are granted full administration rights (e.g., to setup/amend projects), curators are allowed to view and enter data via the web, whereas users can only submit (or receive) data by mobile phone. Any EpiCollect+ project can be flagged as public or private and private projects require a logon to view data. User authentication methods can utilise native logon, Gmail or LDAP and https is supported.\n\n\nDiscussion and future directions\n\nSince its release in 2009, over 4000 EpiCollect v1 projects have been setup at www.epicollect.net using GAE, ranging widely in scope and subject, and v1 continues to be used within a wide variety of disciplines. For example, within Global Health it is being used for OneHealth initiatives12, global wildlife disease monitoring13, vaccination coverage14 and within regional centres of global organisations such as the Food and Agriculture Organisation of the United Nations (FAO)15).\n\nEpiCollect+ substantially increases the flexibility of project design over that available in EpiCollect v1 and provides a generic method for users to define and contribute to complex data gathering projects of many kinds using devices running the Android operating system. EpiCollect+ has been field tested within a number of initiatives to allow the scalability of new functionality to be assessed and refined through feedback from project designers and users. For example, within public health, the Schistosomiasis Consortium for Operational Research and Evaluation (SCORE16), in conjunction with the Schistosomiasis Control Initiative (SCI17) have been using EpiCollect+ for the monitoring and evaluation of praziquantel treatment across multiple African countries. A 5-level hierarchically linked form structure has allowed the collection of data about multiple villages (level 1), multiple locations within villages (schools or houses – level 2) and multiple individuals within each location (level 3). Multiple specimens (blood and stool samples) are taken from each individual and tagged using barcode scanning (level 4) and data are submitted to their central database via mobile phone. Following laboratory processing of samples, data are entered via the web (level 5). One year of data collection in Kenya has resulted in ~250,000 entries being collected by teams of researchers via mobile phone (levels 1–4) and the web (level 5). Instances of the EpiCollect+ server software are setup in-country allowing the complete management of data to/from phones and via the web locally. Further examples include the gathering of human movement data to inform mathematical models of malaria transmission in Mali, Burkina Faso, Zambia and Tanzania (Marshall, J. and Ghani, A., unpublished) and for the assessment of HIV interventions in Zimbabwe (Gregson, S., unpublished).\n\nAnother use of EpiCollect v1, or EpiCollect+ for more complex projects, is within education. Many schools and universities run field courses for students studying different disciplines. For example, the University of Bath (UK) runs an annual field course focussed on basic statistical analysis for students of the biological sciences. Traditionally, students carrying out fieldwork enter data in notebooks and, following transcription onto a computer, the data are collectively analysed to investigate initial hypotheses. Over the past three years, students have undertaken data gathering utilising EpiCollect rather than on paper. Firstly, as data are stored and easily accessible via the web, temporal analysis can easily be undertaken to compare this year’s results with those from previous years, to investigate trends over time. Furthermore, the availability of the project online opens up the possibility of running the same educational project each year in different universities (or schools), and across different countries, increasing the amount of data collected and expanding the scope for data analysis, with the ability to identify both regional and temporal trends.\n\nEpiCollect has also been used for citizen science (e.g., for the collection, by members of the public, of geo-tagged roadkill sightings to understand and inform species protection strategies in the USA18). The increased functionality, and access to media fields in particular, within EpiCollect+ presents further uses within this area. For example, EpiCollect+ has recently been linked with the micro-tasking platform, http://crowdcrafting.org, allowing the processing and interpretation of data gathered via the ‘crowd’. Micro-tasking allows a simple defined task to be undertaken on large sets of data by multiple individuals providing some confidence on the majority result. One EpiCollect+ project gathers recordings of bird song, using audio clips, which are submitted to crowdcrafting for users to specify whether they can identify the particular species of bird. Further citizen science uses of EpiCollect+ are being developed as part of the www.citizencyberlab.eu project.\n\nThe use of EpiCollect v1 or EpiCollect+ should be guided by the requirements of a project. Form building using ‘drag and drop’, or by defining in XML, is more complex in EpiCollect+ than in EpiCollect v1, and this complexity is also apparent when entering data into the mobile app. The simplicity of project setup and use within EpiCollect v1 would make this the best choice for simple projects. The extension of the v1 XML vocabulary allows us to provide backwards compatibility and EpiCollect v1 projects can be loaded into the EpiCollect+ app and we will continue to support both versions. More complex data collection projects will require the much increased functionality of EpiCollect+ and the ability to run projects from the user’s own server. The ability to use ‘drag and drop’ makes the setting up of projects relatively straight forward, and the availability of bundled server software, should allow anyone with moderate knowledge of computing to set up the software and web server to run an EpiCollect+ project in-house.\n\nWe chose to develop EpiCollect+ using Android for a number of reasons. Firstly, Android is an open source platform and has a large community of developers actively contributing to its further development. At the time of the initial EpiCollect publication (2009) only two commercially available handsets running the Android OS were available. Current estimates suggest that there are now almost 400 different handsets (smartphones and tablets) from a variety of hardware manufacturers, ranging widely in price and functionality. Cheap Android devices offering full smartphone functionality are becoming more readily available in resource-poor settings and, in tandem, increases in both mobile data network coverage and speed across the globe are providing higher internet penetration, allowing EpiCollect to be used successfully for projects in most countries.\n\nAt present, EpiCollect and EpiCollect+ are only supported for Android devices. The availability of other mobile operating systems (e.g., Apple iOS, Windows Phone, and Blackberry OS) requires the production of a different codebase for each operating system, increasing development time and costs for each OS supported. A potential solution is provided by HTML5, which is the W3C standard for the production of a platform-independent software that runs entirely within an HTML5 compliant web browser19. A HTML5 version of EpiCollect+, available for multiple operating systems, is currently in BETA testing.\n\n\nSoftware availability\n\nThe EpiCollect+ Android app can be downloaded from Google Play.\n\nThe EpiCollect+ Server and instructions are available at: http://www.epicollect.net.\n\nThe server: https://github.com/ImperialCollegeLondon/EpiCollectplus\n\nThe mobile app: https://github.com/ImperialCollegeLondon/EcPlusAndroid\n\nhttp://dx.doi.org/10.5281/zenodo.1127820\n\nhttp://dx.doi.org/10.5281/zenodo.1128121\n\nAGPLv3 Licence", "appendix": "Author contributions\n\n\n\nConceived the study/application: DMA, DMH, BGS. Server coding and epicollect.net: CIP, DMA. Android Coding: DMH. HTML5 Coding: MM. Wrote the paper: DMA, BGS.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe work was funded by Wellcome Trust grant WT089472, BGS is a Wellcome Trust Principal Research Fellow, CP was funded by UKCRC Grant (MRC G1000803), MM is funded by EU FP7 Grant: CitizenCyberLab (FP317705).\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 acknowledge a number of field workers who have tested EpiCollect+ for their data collection projects and have provided useful feedback.\n\n\nReferences\n\nThe Android Operating System. Reference Source\n\nGSM Roaming and Coverage Maps. Reference Source\n\nRapidSMS – SMS Application Framework. Reference Source\n\nEpisurveyor. Reference Source\n\nHartung C, Ankowa Y, Brunette W, et al.: Open data kit: tools to build information services for developing regions. ICTD2010; Surrey, UK: 2010. Publisher Full Text\n\nAanensen DM, Huntley DM, Feil EJ, et al.: EpiCollect: linking smartphones to web applications for epidemiology, ecology and community data collection. PLoS One. 2009; 4(9): e6968. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoogle App Engine. Reference Source\n\nW3C XML Specification. Reference Source\n\nArduino, Open Source electronics prototyping platform. Reference Source\n\nZxing, Open Source Barcode Scanning Library. Reference Source\n\nXAMPP – Open Source Apache Installation containing MySQL, PHP and Perl. Reference Source\n\nRweyemamu MM, Mmbuji P, Karimuribo E, et al.: The Southern African Centre for infectious disease surveillance: a one health consortium. Emerg Health Threats J. 2013; 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOlson DH, Aanensen DM, Ronnenberg KL, et al.: Mapping the global emergence of Batrachochytrium dendrobatidis, the amphibian chytrid fungus. PLoS One. 2013; 8(2): e56802. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThe New Indian Express: Mission Rabies or mission Impossible? Reference Source\n\nFood and Agriculture Organization of the United Nations (FAO) 2013 – Cell phones revolutionizing Kenya’s livestock sector. Reference Source\n\nThe Schistosomiasis Consortium for Operational Research and Evaluation. Reference Source\n\nThe Schistosomiasis Control Initiative. Reference Source\n\nNational Public Radio Dec 2012 – The sight of Roadkill makes a pretty Data-Rich picture. Reference Source\n\nW3C HTML5 Specification. Reference Source\n\nAanensen DM, Huntley DM, Menegazzo M, et al.: F1000Research/EpiCollectplus. Zenodo. 2014. Data Source\n\nAanensen DM, Huntley DM, Menegazzo M, et al.: F1000Research/EcPlusAndroid. Zenodo. 2014. Data Source" }	[ { "id": "6371", "date": "22 Oct 2014", "name": "Daniel Lombraña González", "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 describes properly the development and features of the software EpiCollect and its new version EpiCollect+.The paper describes in detail how the software works, giving a proper step by step tutorial about how to create a project, use the project, and see the charts and maps related to the collected data.As a suggestion for improving the paper, I would say that:A table comparison between different frameworks/solutions in order to have a clear idea about what makes EpiCollect+ different from other solutions. The paper should address a bit more of the new features of its new HTML5 version, giving a bit more of information about its beta version.In general I recommend this paper to be accepted and indexed.", "responses": [] }, { "id": "5899", "date": "17 Nov 2014", "name": "Jason E. Stajich", "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 provides a description of the EpiCollect+ tool for smartphone based collection of data for web applications. It describes open source software for the Android platform to build up data collection sources and integration into databases.", "responses": [] } ]	1	https://f1000research.com/articles/3-199
https://f1000research.com/articles/3-198/v1	19 Aug 14	{ "type": "Review", "title": "Current treatment of early breast cancer: adjuvant and neoadjuvant therapy", "authors": [ "Elizabeth Miller", "Hee Jin Lee", "Amriti Lulla", "Liz Hernandez", "Prashanth Gokare", "Bora Lim", "Elizabeth Miller", "Hee Jin Lee", "Amriti Lulla", "Liz Hernandez", "Prashanth Gokare" ], "abstract": "Breast cancer is the most commonly diagnosed cancer in women. The latest world cancer statistics calculated by the International Agency for Research on Cancer (IARC) revealed that 1,677,000 women were diagnosed with breast cancer in 2012 and 577,000 died. The TNM classification of malignant tumor (TNM) is the most commonly used staging system for breast cancer. Breast cancer is a group of very heterogeneous diseases. The molecular subtype of breast cancer carries important predictive and prognostic values, and thus has been incorporated in the basic initial process of breast cancer assessment/diagnosis. Molecular subtypes of breast cancers are divided into human epidermal growth factor receptor 2 positive (HER2 +), hormone receptor positive (estrogen or progesterone +), both positive, and triple negative breast cancer. By virtue of early detection via mammogram, the majority of breast cancers in developed parts of world are diagnosed in the early stage of the disease. Early stage breast cancers can be completely resected by surgery. Over time however, the disease may come back even after complete resection, which has prompted the development of an adjuvant therapy. Surgery followed by adjuvant treatment has been the gold standard for breast cancer treatment for a long time. More recently, neoadjuvant treatment has been recognized as an important strategy in biomarker and target evaluation. It is clinically indicated for patients with large tumor size, high nodal involvement, an inflammatory component, or for those wish to preserve remnant breast tissue. Here we review the most up to date conventional and developing treatments for different subtypes of early stage breast cancer.", "keywords": [ "adjuvant treatment of breast cancer", "breast cancer", "current treatment", "HER2 targeted therapy", "hormone receptor positive breast cancer", "neoadjuvant treatment" ], "content": "Introduction\n\nBreast cancer is the most commonly diagnosed cancer in women. The latest world cancer statistics available from the International Agency for Research on Cancer (IARC) showed that 1,677,000 women were diagnosed with breast cancer and 577,000 women died in 20121. Improvements in chemotherapy, surgery, lymph node evaluation and hormone receptor blocking therapy have successfully doubled the survival of breast cancer patients2. The evolution of genomic research enabled the genetic and molecular profiling of cancers, which also revealed the profound complexity and heterogeneity of breast cancer3–5. Different molecular subtypes of breast cancer have various prognoses and responses to therapy6. Such complexity makes it challenging for clinicians to keep abreast of new knowledge and novel. Therefore, this review gives an overview of current treatments for breast cancer. We will review treatment options based on the different stages and the molecular subtypes of breast cancer that are commonly used in the United States and Europe.\n\n\nAdjuvant treatment in hormone receptor positive breast cancer\n\nHormone receptor (estrogen and progesterone) positive breast cancers account for the largest portion of diagnosed breast cancers. The hormone receptor positive breast cancers constitute up to 65–75% of all breast cancers and this proportion is rising7. The cells of this subtype of breast cancer are largely dependent on female hormone supply for their growth and survival8. The understanding of the related biology is important in treatment design. Breast cancers that express hormone receptors (either estrogen or progesterone), but not the human epidermal growth factor receptor 2 (HER2) protein, are categorized as luminal A intrinsic subtype4. Ki67, a nuclear protein that is encoded by MIKI67 gene is a marker of proliferation, and is also critical in differentiating between A and B luminal subtypes9. The luminal A subtype of breast cancer has the best prognosis amongst all subtypes, but even so, up to 20% of early stage luminal. A breast cancer patients experience breast cancer recurrence within 10 years after the completion of initial treatment without adjuvant treatment4. Two main adjuvant therapy modalities are cytotoxic chemotherapy and endocrine (hormone receptor blocking) therapy. Both adjuvant treatment modalities improve disease free survival (DFS) and overall survival (OS) in hormone receptor positive breast cancer patients10.\n\nThe Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) compared an anthracycline-based regimen with a CMF (cyclophosphamide, methotrexate, and fluorouracil) regimen that was used more commonly in breast cancer starting from 197011. In 2001, EBCTCG reported the collective data of the randomized trials in early breast cancer adjuvant systemic chemotherapy from 1985 to 200012. This report not only showed the long term benefits of an adjuvant endocrine therapy, but also confirmed a 50% reduction of the overall mortality in 15 years, when hormone receptor positive breast cancer patients received adjuvant chemotherapy and tamoxifen for 5 years following surgery12. EBCTCG subsequently reported the 10 year follow-up results after the initial report: when compared to the untreated group, the anthracycline-based chemotherapy group had an absolute gain of 8% in recurrence free survival, 6.5% in breast cancer mortality, and 5% in overall mortality. The CMF regimen similarly improved survival, but achieved 10.2% of absolute gain of recurrence free survival13. CMF is one of the oldest poly-chemotherapy regimens developed for breast cancer. It was first introduced by Bonadonna et al.15. CMF was initially given every month for a total of 12 months after primary breast surgery. Later, the same group compared 6 cycles versus 12 cycles of CMF, which showed no difference in both relapse free/OS in the two groups. An EBCTCG meta-analysis calculated a reduction of 6.2% in absolute breast cancer related mortality at 10 years follow-up, when using CMF adjuvant therapy, compared to no adjuvant chemotherapy16.\n\nAnthracycline was the next important agent to improve the efficacy of adjuvant chemotherapy. The NSABP B-11, B-12 trials showed the efficacy of doxorubicin (Adriamycin) in stage II breast cancer patients. The group who received doxorubicin plus melphalan and fluorouracil (PF) had significantly improved DFS and OS in 6 years, compared to those who received the same regimen without doxorubicin17. NSABP B-15 compared 4 cycles of AC (Adriamycin 60mg/m2 plus cyclophosphamide 600mg/m2) with the conventional 6 cycles of CMF. A total of 2194 patients with positive nodes and a negative estrogen receptor were randomized into these two groups, and no difference in DFS or OS were shown between the two groups at the 10 year follow-up18. The NSABP B-23 trial compared node negative, estrogen receptor negative patients randomized to 4 cycles of AC versus CMF, and again showed the same DFS and OS in both arms19. NSABP B-16 compared 4 cycles of AC plus tamoxifen and tamoxifen alone as adjuvant therapy. This trial showed a 15% proportional reduction in average annual hazard, relapse, or death at 10 years follow-up, and a 25% relative risk reduction in comparison of tamoxifen only. Based on an indirect comparison, the degree of risk reduction in overall breast cancer related morbidity and mortality of anthracycline based regimen was much greater than that of CMF, which was 10%20. The CALGB49907 trial was a randomized trial comparing CMF or AC chemotherapies with capecitabine as adjuvant chemotherapy for patients older than 65 with non-metastatic (stage I to IIIB) breast cancer. In this trial, the patients who received chemotherapy (either CMF or AC) had better relapse free survival at 3 years follow-up, resulted in 85% in chemotherapy group versus 68% in capecitabine arm respectively21. A meta-analytical comparison of CMF versus anthracycline containing poly-chemotherapy regimen in the adjuvant setting showed that the standard 4 cycles of AC and the standard CMF were equivalent (RR 0.98, SE 0.05, 2p=0.67), but anthracycline-based regimens with substantially higher cumulative dosage than the initial standard regimen of 4 AC cycles (e.g., CAF or CEF) were significantly superior to the standard CMF13. This offered the rationale for the development of an anthracycline based poly-chemotherapy in an adjuvant setting.\n\nThe second biggest game changer in breast cancer adjuvant chemotherapy was the introduction of taxane. After the efficacy of taxane was shown in advanced breast cancer, the BIG 02-98 trial incorporated docetaxel into the adjuvant setting. This trial compared sequential versus concurrent doxorubicin and docetaxel chemotherapy for lymph node positive breast cancer, showing docetaxel arms with improved survivals22. BCIRG001 is an open label phase III multicenter randomized trial comparing early breast cancer with positive nodes who received TAC (docetaxel, doxorubicin, and cyclophosphamide), or FAC (fluorouracil, doxorubicin, and cyclophosphamide) 3 times a week for 6 cycles. Primary end point of the study was DFS. At a median follow-up of 10 years, TAC group had better DFS and OS compared to FAC group. Improved DFS in TAC group was not dependent on nodal status, hormone receptor or HER2 status23. The Intergroup 9344 (INT 9344) trial that was led by NSABP in collaboration with the Eastern Cooperative Oncology Group (ECOG) and South Western Oncology Group (SWOG), that to address the question of whether adding 4 cycles of paclitaxel (T) to 4 cycles of AC would improve the clinical outcome. There was a 5% absolute improvement in DFS and a 3% in OS by adding paclitaxel (T), but not by adding cycles of adriamycin24. The NSABP B-28 study compared 4 cycles of AC versus 4 AC plus 4 T. This trial also showed that adding T resulted in relative DFS improvement of 17%, but lesser degree of improvement in OS (7%)25. Based on the result from the EBCTCG meta-analysis, the overall absolute reduction of recurrence by additional taxane to the anthracycline regimen was 2.8% and the reduction of mortality with recurrence was 1.3%. However this improvement was diluted when a very well dosed anthracycline-based regimen was used. The improvement in clinical outcome was sustained over a period of 5 years13.\n\nNot only the selection of chemotherapy agents but also the method of delivery is critical in the development of chemotherapy. The CALGB 9741 study compared 4 arms with different dosing schedules of AC-T. The first two arms were given a total of 4 cycles of all regimens every 3 weeks, while the second two arms received treatment every 2 weeks. A protocol-specified analysis was performed at a median follow-up of 36 months. Q 2 weekly dose-dense (dd) schedule improved the DFS and OS. However, there was no difference in either DFS or OS between the dd concurrent and the dd sequential schedule arms26. Dose dense schedule is widely used as an AC-T schedule unless there are other factors. More recently, SWOG S0221 conducted a 2×2 design phase III trial comparing AC+G (filgrastim) versus dd AC in different combinations with either dd T (paclitaxel) for 6 cycles or weekly T for a total of 12 weeks. This trial showed equivalent progression free survival (PFS) in both weekly and 2 weekly T (82% in weekly versus 81% in dd PFS), suggesting that weekly paclitaxel could have the same efficacy without the patients having to receive the growth factor support27. A Spanish group published the results from the GEICAM/2003-02 study, comparing FAC, and FAC followed by weekly paclitaxel for node negative high-risk patients. In this study, additional weekly paclitaxel for 8 weeks added a 2.7% improvement in PFS at 63.3 months follow-up28.\n\nThere are two main categories of endocrine therapy agents: selective estrogen receptor modulators (SERMs) and aromatase inhibitors (AIs). SERMs competitively bind to estrogen receptors to interfere with DNA synthesis by recruiting co-repressors, and inhibit G0->G1 cell cycle progression29. The three main drugs of this category are tamoxifen, raloxifen, and toremifene. AIs work differently. These drugs inhibit an enzyme called ‘aromatase’ that converts circulating testosterone to estradiol (E2), and androstenedione to estrone, by aromatization. Such peripheral conversion of other hormones to estradiol is the main source of estrogen in post-menopausal women30. Therefore, AIs only work when the primary source of estrogen is terminated – either by the menopausal state, oophorectomy, or estrogen deprivation therapy using luteinizing-hormone-releasing hormone (LHRH) agonists. Exemestane, anastrazole and letrozole are three main drugs of this category.\n\nTamoxifen and its effects have been studied for over 3 decades in thousands of women, as a primary and secondary preventive therapeutics. It is estimated that 400,000 or more women are estimated to be alive as a result of tamoxifen therapy worldwide and that also due to tamoxifen, millions of women achieved extended DFS31,32. Fifteen years of adjuvant treatment review of EBCTCG concluded that tamoxifen successfully reduced the absolute rate of breast cancer recurrence in hormone receptor positive early stage breast cancer by 13% (2p<0.00001), and breast cancer related mortality by 9.1% (2p<0.00001)12. A 5 year duration of adjuvant tamoxifen has been the standard of care for years, however ATLAS (Adjuvant Tamoxifen: Longer Against Shorter) trial showed the benefit of longer tamoxifen use. This trial accrued 80,000 women and randomized them to extend the tamoxifen therapy for 10 years versus stopping at 5 years as previously recommended. The extended treatment arm to 10 years had a 4% improvement in breast cancer related mortality33. Since the result of this study, extended duration of endocrine therapy has been incorporated as the standard of care, as long as the patient can tolerate the treatment without side effects.\n\nIn post-menopausal women, AI is the regimen of choice, based on the improved efficacy compared to tamoxifen as shown in previous trials. The ATAC (anastrazole, tamoxifen, alone or in combination) trial compared the efficacy of anastrazole and tamoxifen for postmenopausal women in adjuvant settings. After a median follow-up of 68 months, anastrazole showed a significantly prolonged DFS compared to tamoxifen, significantly reduced distant metastases (324 vs 375; HR 0.86 and contralateral breast cancers (35 vs 59; 42% reduction)34.\n\nDuring the 5–10 years of endocrine therapy, a patient’s menstrual status can change from premenopausal to a menopausal state. Therefore, it is not surprising to raise the question whether the use of different endocrine adjuvant therapies in sequence could affect the clinical outcome. The MA-17 trial enrolled 5170 post-menopausal patients who had completed 5 years of adjuvant tamoxifen, and assigned them either to receive an additional 5 years of letrozole or a placebo. The DFS at 4 years follow up was 94.4% in the letrozole arm versus 89.8% in the placebo arm – representing 4.6% of absolute reduction in disease recurrence. Both distant recurrence and contralateral breast cancer incidence were lower in the additional letrozole adjuvant arm35. The BIG-98 trial compared three groups – one group received letrozole for 5 years, another one received tamoxifen for 5 years, and the last group received sequential therapy. The letrozole arms were superior, but the DFS and OS of the sequential therapy were the same as using letrozole monotherapy. The outcome of tamoxifen followed by a letrozole arm was the same as for the letrozole monotherapy, but there was a trend towards a better outcome in the letrozole monotherapy arm, suggesting the superiority of letrozole as a first line endocrine therapy36. Dowsett et al. compared two cohorts of postmenopausal patients’ data by meta-analysis. Cohort 1 patients started endocrine therapy with AI and continued to take AI, or converted therapy from tamoxifen to AI. At 5 years, the AI monotherapy resulted in an absolute 2.9% reduction in recurrence (9.6% for AI versus 12.6% for tamoxifen; 2P < .00001) and a non-significant 1.1% absolute reduction in breast cancer mortality (4.8% for AI vs 5.9% for tamoxifen; 2P = 0.1). Cohort 2 patients started endocrine therapy with tamoxifen for 2 years, then they were randomized to either continue tamoxifen or switch to AI. At 3 years from treatment divergence (which was about 5 years after the initiation of endocrine therapy), the group who converted the therapy to AI showed an absolute 3.1% recurrence and an absolute 0.7% reduction in breast cancer mortality37. From these data taken together, AI is considered as the gold standard first line therapy for post-menopausal women in adjuvant endocrine therapy.\n\nDespite the proven benefits of chemotherapy in early stage hormone receptor positive breast cancer patients, it is also clear that the absolute benefits of chemotherapy are not the same across all patients. Traditionally, gender, ethnicity, pathologic stage of tumor, age, personal history and family history were considered to be the main factors that could help to measure the benefit of adjuvant chemotherapy in individual patients38. Adjuvant Online!39, a commonly used risk calculator, is a good example of such a traditional measure of prognosis. It has been widely studied and validated in different populations of patients12,40,41. Over time, we have learned that the biologic characteristics of tumors can be more critical in adjuvant treatment decision making. Several comprehensive genomic profiling tools to characterize and predict the prognosis of individual patients have been developed. Such genomic profiling tools not only provide sub-typing of breast cancers, but also can predict their response to adjuvant therapy. For instance, Oncotype DX™ calculates the prognosis of individual patient’s 10 years recurrence risk by assessing 16 genes that are related to the proliferation of the tumor. Intriguingly, this tool also gives a validated prediction as to whether the individual patient who receives adjuvant endocrine treatment tamoxifen will have an additional benefit by the addition of chemotherapy. The predictive value of Oncotype DX™ was validated in both pre- and post-menopausal women42,43. However, there were questions remaining for patients with an intermediate score from Oncotype DX™ recurrence score testing. TAILORx (the Trial Assigning IndividuaLized Options for Treatment : NCT00310180) is being conducted to answer this question44.\n\nThe application of recurrence score has been expanded to node positive patients as well. A retrospective ancillary study that analyzed tumor samples from the SWOG-8814 trial node positive breast cancer patients confirmed that patients with low Oncotype DX™ did not gain additional benefits from chemotherapy , also in some patients with positive lymph nodes. For patients with a high recurrence score on the other hand, there was significant improvement of progression free survival independent of the number of positive nodes (hazard ratio 0.59; 95% CI=0.35–1.01)45. SWOG S-1007 RxPONDER (Rx for Positive Node, Endocrine Responsive Breast Cancer)46, a prospective, randomized trial is currently ongoing to further determine the effect of chemotherapy for patients with up to 3 positive lymph nodes involvement. PAM50 is a more comprehensive genomic profiling tool for breast cancer. This test not only can detect the intrinsic subtype of breast cancer, but also will predict the prognosis of individual patients47. However, so far the studies have not been able to validate the predictive value of PAM50 for specific therapeutic use48, thus, to date, it is mainly used for prognosis/sub-typing reasons. High risk diseases that were detected via the PAM50 test are mainly non-luminal A or B cancers, the majority of them being triple negative breast cancers. MammaPrint49 is another available genetic signature mainly currently used in Europe using microarray chip technology. This tool analyzes a total of 70 genes that represent 6 hallmarks of cancer. The test is also currently developed as a predictive marker for better selection of tailored therapies for breast cancer patients.\n\nTraditionally, bisphosphonate was used to treat hypercalcemia and osteoporosis as it blocks the activity of osteoclasts. Because of the relationship between cancer cells and osteoclasts in the bone marrow niche acting as a feedback loop in an interconnected microenvironment, bisphosphonate also has a great activity against bony metastasis in solid cancers. Moreover, osteoclast secreted RANKL (receptor activator of nuclear factor kappa-B ligand) and RANK combination promotes the proliferation and survival of breast cancer stem cells in pre-clinical studies, suggesting a strong scientific rationale to use inhibitors of the osteoclast activity to improve survival50. However, the results from various small trials were mixed, and resulted in confusion and debate in the field. The AZURE study was the first trial to bring interest and attention towards bisphosphonate in the field by showing the survival benefits. The OS of the zolendronic acid adjuvant treatment group and a control group were 85.4% versus 83.1% respectively, with a confidence interval of 151. NSABP B-3452, GAIN53 and NATAN54 trials did not show the same superiority in an adjuvant bisphosphonate use group. The actual agents used in the different trials varied – both IV and oral agents were used. Interestingly, when subgroup analysis was performed either by age 55 or menopausal status, there was improved hazard ratio in DFS in elderly, post-menopausal women. The p-values in later 3 trials were not statistically significant. Recently, large meta-analysis done by a group in United Kingdom collectively analyzed total of 18000 women from 41 different studies55. The results among pre-menopausal women did not show any difference between the bisphosphonate group and non-bisphosphonate group in both recurrence free survival and breast cancer related mortality. However in menopausal women, the breast cancer related mortality was reduced by 3.1%, and the distant recurrence rate was reduced by 3.5%. Given the low side effect profile of the drug, this result will likely change standard practice in the near future.\n\n\nAdjuvant therapy in HER2 positive breast cancer\n\nAbout 20–25% of breast cancers are characterized by the over-expression of HER2 protein56. HER2 (ErbB2) is a transmembrane glycoprotein that has both an intracellular receptor tyrosine kinase (TK) domain and an extracellular ligand binding domain. The HER (ErbB) family consists of HER1 (ErbB1 = EGFR), 2, 3, and 457. Different subtypes of HER protein share similar intracellular TK domains, but express distinct ligand binding extracellular domains56. HER receptors are activated via homodimerization, or heterodimerization with its family member HER1 and HER3. HER2 overexpression is one of the most important carcinogenic features, as well as being a prognostic and predictive marker for response to HER2 targeted therapy56. Trastuzumab is the first monoclonal antibody developed as an anti-HER2 therapeutic that binds to the juxtamembrane domain of HER2 receptor58. Trastuzumab has other interesting activities; it induces the activity of p21 or p27, which then cause transcription inhibition and also induces antibody-dependent cell – mediated cytotoxicity (ADCC)59. Since the first US Food and Drug Administration approval in 1997, trastuzumab has become a cornerstone of HER2 overexpressing breast cancer treatment in any stage of disease, including the adjuvant setting.\n\nThe BCIRG 006 trial accrued early stage HER2 overexpressing breast cancer patients between April 2001 and March 2004 and compared three arms: AC-T (adriamycin, cyclophosphamide, and paclitaxel), AC-TH (adriamycin, cyclophosphamide, paclitaxel and trastuzumab), and TCH (docetaxel, cyclophosphamide, and trastuzumab). The primary endpoint of DFS after a median follow-up of 65 months was 75% in the AC-T arm, 84% in the AC-TH arm, and 81% in the TCH arm. The first planned interim analysis was performed in 2006. TCH had a better side effect profile, and without a non-statistically significant difference in efficacy this led to its approval by the FDA60. NCCTG (North Central Cancer Treatment Group) N9831, NSABP (National Surgical Adjuvant Breast and Bowel Project) B-31, FinHER (Finland Herceptin), HERA, NOAH (Neo-adjuvant Herceptin), FNCLCC-PACS (Federation Nationale des Centres de Lutte Contre le Cancer-Programmes d’Actions Concertees Sein) 04, BCIRG (Breast Cancer International Research Group) 006 trials all showed that a trastuzumab – chemotherapy combination regimen - leads to improved clinical outcome compared to conventional cytotoxic adjuvant therapy61–65.\n\nThe next question to address is the duration of adjuvant treatment. The HERA trial was an open label, large randomized phase III trial comparing 2 years versus 1 year use of adjuvant trastuzumab for patients with HER2 positive breast cancer. A total of 5102 patients were randomized into two groups, after completion of 1 year adjuvant trastuzumab to either stop at year point, versus 1 additional year to complete 2 years. The primary end point of this study was a PFS, and there was no difference between two groups62,66. Therefore, a year (52 weeks) of adjuvant trastuzumab treatment after surgery is currently the standard of care for early stage HER2 positive breast cancers.\n\nHowever, resistance to trastuzumab therapy still remains a challenge in the treatment of HER2 overexpressing breast cancer. HER1 or HER3 can bind to the ligand, and can activate the intracellular downstream signaling of cancer cells regardless of HER2 blockage therapy. The other common mechanisms of resistance to trastuzumab include the truncated form of HER2. If the HER2 protein lacks the antibody binding domain (the truncated form of HER2 is also called p95)67, it is resistant to trastuzumab due to lack of an appropriate binding site. The phosphatase and tensin homolog (PTEN) gene mutation, resulting in PTEN constant activation, can bypass the blockage of HER2-mediated intracellular signaling68, insulin-like growth factor 1 receptor (IGF-R), and phosphoinositide 3-kinase (PI3KA)/Akt pathway amplification69. Defective apoptosis pathways are main causes of trastuzumab resistance. Most recently, immunologic factors - different expression of stromal tumor infiltrating lymphocytes70, defective Fc receptors71 that can interfere with normal immune responses to trastuzumab have also been suggested as mechanisms of resistance.\n\nLapatinib was the first small molecule that was developed to overcome trastuzumab resistance. Lapatinib not only inhibits HER2 but also inhibits HER1 (=EGFR), although a later preclincal study72 suggested that the activity of lapatinib was HER1 independent. It binds to the intracellular domain of HER2 protein, thus the efficacy is preserved for the truncated form of HER2 protein. This drug shows efficacy as a single agent, in combination with capecitabine, and with trastuzumab in metastatic settings73. Unfortunately, lapatinib failed to show the efficacy in an adjuvant setting. A total of 3161 women who may have received adjuvant therapy without trastuzumab were divided into lapatinib and placebo group. DFS at 47.4 months follow up showed 13% in lapatinib versus 17% in placebo group. However this study included patients who had no HER2 protein expression by central review. The authors reported that lapatinib had a marginal benefit in women with confirmed HER2 positive breast cancer, but this may suggest inconclusive benefits of lapatinib in adjuvant settings74. TD-M (emtansine-trastuzumab conjugate), and pertuzumab are newer agents targeting HER2 overexpression in breast cancer, and currently approved for use in metastatic settings by the US FDA75. Pertuzumab has also been approved in neoadjuvant settings. These two agents will be discussed in the sections below.\n\n\nAdjuvant therapy for triple negative breast cancer\n\nTriple negative breast cancer (TNBC) is a subtype of breast cancer that accounts for 10–15% of breast cancer cases. TNBC is a heterogeneous group of tumors that commonly occur in younger women, African Americans, and in BRCA gene-mutated populations76,77. It is called ‘triple negative’ because this subtype of breast cancers are negative for ER/PR/HER276. The survival of patients with metastatic or recurrent TNBC remains poor to date, due to lack of meaningful biologic targets, and the recurrence rate is higher than other subtypes of breast cancers when compared at same stage of disease78. The benefit of an adjuvant chemotherapy in TNBC is greater than in hormone receptor positive breast cancer patients, based on the data from a large meta-analysis by EBCTCG79. Given the lack of effective targets in this subtype of breast cancers, chemotherapy remains as mainstay of adjuvant therapy for TNBC.\n\nStandard regimens currently used in TNBC subgroup are the same as for hormone receptor positive cancers, since this subgroup of tumors responds well to both anthracycline or taxane based regimen80,81. The benefit of an adjuvant cytotoxic therapy is much greater in TNBC. A retrospective analysis of three large CALGB trials including 6,444 patients confirmed the substantially larger benefits of adjuvant chemotherapies for hormone receptor negative breast cancers. When comparing two different chemotherapy regimens as adjuvant treatments – CAF (cyclophosphamide, adriamycin, 5-FU) with dose dense Q2 weekly AC-T from different CALGB trials, there were a 55% relative reduction and a 28% absolute reduction of recurrent risk for hormone receptor negative tumors26. Thus, TNBC patients with a tumor size greater than 1cm, or any lymph node involvement, receive cytotoxic chemotherapy unless they have significant contraindications.\n\nOne sub-group of patients among the patients with TNBC, treated with a more promising targeted therapy currently under development, are the patients with BRCA1 and 2 mutations. BRCA1 and 2 are important DNA repair genes, thus the tumorigenesis in this proportion of TNBC subtypes is higher in this population. About 10% of women with TNBC who had a cancer diagnosis at less than age 40 are found to have BRCA1 or 2 mutations82. Women with TNBC younger than 50 years old could present up to 10–25% BRCA1/2 mutation incidence83, which offers a good rationale for using platinum agents. In addition, more than half of TNBCs have mutation of TP5347, which gives another rationale of platinum sensitivity, given the pre-clinical data from breast cancer cell studies showing that cells are more sensitive to platinum agents when they have a defect/mutation in the p53 family proteins. Currently, the platinum agents – mainly carboplatin – are studied in neoadjuvant settings where the study outcome can be assessed in a short period of time. However the use of platinum agents in adjuvant setting is continuously evolving in this subgroup of breast cancers.\n\nPre-clinical studies and early phase clinical studies revealed the importance of angiogenesis and microenvironment in triple negative breast cancer cells, suggesting the efficacy of VEGF targeted therapy84,85. Based on exciting early data, the BEATRICE trial enrolled total 2591 patients with early stage breast cancer, divided them into two groups – one group to receive standard adjuvant chemotherapy and monitor, and the other to receive standard chemotherapy (either anthracycline or taxane based on investigator’s choice) + 5mg/kg weekly equivalent bevacizumab, and followed by bevacizumab maintenance. Unfortunately, 3-year DFS was 82.7% in the chemotherapy only group versus 83.7% in the bevacizumab group. There was a certain trend towards bevacizumab benefits in patients who had a high pre-treatment plasma VEGFR-2 level, but this was not statistically significant, disappointing researchers and clinicians in the field86.\n\n\nNeoadjuvant therapy\n\nTraditional indications for neoadjuvant therapy in breast cancer include N2 stage - fixed or matted lymph node on ipsilateral side, or clinically apparent ipsilateral internal mammary nodes in the absence of axillary node, making the clinical staging at least stage IIIA or above. Patients with stage IIIB disease with tumors invading the chest wall, skin or both, or with breast cancer of inflammatory nature, would be a good candidate for neoadjuvant therapy87. Neoadjuvant therapy should also be considered for women with clinical stage IIA and IIB tumors with a larger tumor who wish to have breast-conserving operations and avoid mastectomy. Not in all, but in many patients, neoadjuvant therapy results in sufficient tumor response to make breast-conserving operations possible. Several studies in the early 2000s showed that neoadjuvant chemotherapy successfully reduced both locoregional and in breast tumor recurrence even in large T3 and T4 tumors88,89. More recently, pathologic complete remission (pCR) is an important concept currently developed as a prognostic marker of survival in breast cancer patients that can be used as a surrogate outcome of survival53,90. Neoadjuvant therapy has been evolving rapidly given this benefit91.\n\n\nNeoadjuvant therapy for hormone receptor positive breast cancer\n\nThere have been many trials comparing the clinical outcome of pre-operative versus post-operative therapy. The EORTC10902 trial accrued 698 patients early stage breast cancer (both hormone receptor positive and negative) randomized to 4 cycles of 5-FU, epirubicin, and cyclophosphamide (FEC) administered pre-operatively versus the same regimen given post-operatively. PFS, OS, or local recurrence rate were not different when comparing pre-operative and post-operative therapy92.\n\nFrom the meta-analysis, Mauri and colleagues found no difference with regard to death (RR 1.00, 95% CI, 0.90 to 1.12), disease progression (RR 0.99, 95% CI, 0.91 to 1.07), or distant disease recurrence (RR 0.94, 95% CI, 0.83 to 1.06). However, the rate of local recurrence was higher in the neoadjuvant group (RR 1.22, 95% CI, 1.04 to 1.43). This was mainly in trials where surgery was avoided in cases of clinical complete response93.\n\nNonetheless, for an individual patient, the delay of surgery by pre-operative therapy could provide potential harm. Given that all randomized trials are comparisons of cohorts, the disadvantages of single patients are not reflected in the overall results. Fortunately, the proportion of tumors progressing during neoadjuvant therapy is very low, but hypothetically even if the tumor as a whole is shrinking, single tumor cells could respond differentially. As discussed previously, partly resistant tumor cells might acquire full-blown resistance during neoadjuvant treatment and generate micrometastases.\n\nA broad use of endocrine therapy as a tool in neoadjuvant settings could be somewhat limited due to the slow response rate of tumors in general, requiring long duration of therapy and risking the benefit of early surgical intervention94. Also, other important benefits of using a neoadjuvant therapy – to assess the response of tumor to the treatment, to explore the prediction of long term relapse free survival - are less obvious in hormone receptor positive breast cancers91. Thus, an endocrine therapy as a neoadjuvant therapy tool has been tested primarily in postmenopausal women who aimed to change the extent of surgical interventions from a mastectomy to a breast-conserving operation, but who were not fit for chemotherapy due to medical co-morbidities. When the response rates were compared between AI and tamoxifen in this setting, the clinical response rate was significantly higher in the AI group than in the tamoxifen group, but overall the pCR was less than 10%95. To date, there has not been a direct comparison of long term progression free, or overall survival between neoadjuvant endocrine therapy followed by surgery versus surgery followed by adjuvant therapy.\n\nA study conducted by Cameron et al. compared endocrine neoadjuvant therapy only group with a group who received both endocrine therapy and cytotoxic therapy, after being found not to have significant response in operable breast cancer patients, and two groups did not show a difference in clinical outcome96. The number of axillary lymph nodes after systemic chemotherapy, and the rate of response after either endocrine or pre-operative chemotherapy were the factors predicting survival outcome, the number of lymph nodes more strongly so, suggesting that as long as the patient responds to hormone targeted therapy, the prognosis is good despite the overall low response rate97.\n\nWhen Semiglazov et al. compared anastrozole to chemotherapy in elderly women in, both had equivalent benefit when used in a neoadjuvant setting for women older than 70, with hormone receptor positive cancer98. In the IMPACT trial which compared anastrozole, tamoxifen, and a combination of both agents in postmenopausal women, three arms showed similar response rate. However, for women who needed mastectomy at baseline, anastrozole showed a significant improvement in terms of downgrading the extent of surgery from a mastectomy to a breast-conserving operation99. There is currently not much information available on the use of neoadjuvant endocrine treatment in premenopausal patients.\n\nA phase II study was conducted to see whether everolimus added to letrozole for operable breast cancer patients would improve the clinical response. Everolimus is an inhibitor of mTOR, a downstream signaling molecule of Akt. A group who received the everolimus combination had higher response compared to the letrozole and placebo group (68% versus 59.1%). The everolimus combination arm also showed a reduction in the expression of the biomarker Ki67, as well as a phosphor-S6, in surgical specimens, suggesting downregulation of the proliferation and PI3K pathway. The safety profile of the everolimus arm was similar to monotherapy with everolimus100. These results suggest that an including an additional mTOR inhibitor with the hormone therapy may improve the overall response rate in neoadjuvant settings. However the potential benefit on survival rate is still unclear, hence it needs to be further investigated.\n\nDespite the lower response rate of hormone receptor positive breast cancer patients to chemotherapy, compared to patients with other subtypes of cancer, chemotherapy is still the main neoadjuvant systemic therapy that has been widely studied and used101,102. So far, molecular profiling to predict the benefits of using different options of therapy in the neoadjuvant setting have not been well studied. Therefore, there is no good prediction tool to select the perfect candidate of neoadjuvant chemotherapy among hormone receptor positive breast cancer patients. Because of this reason, in the treatment of operable hormone receptor positive breast cancer patients, a more personalized neoadjuvant systemic cytotoxic therapy based on the patient’s wish and clinical scenario is preferred rather than a standardized therapy.\n\nAnthracycline/taxane-based chemotherapy regimens have been studied extensively in prospective randomized trials. Overall, pCR is between 15% and 20% in hormone receptor positive breast cancer patients who are pre-operatively treated with cytotoxic therapy. The main regimens studied in neoadjuvant settings include AC followed by docetaxel or paclitaxel, epirubicin/paclitaxel-CMF, and a dose-dense sequence of epirubicin and paclitaxel103–106.\n\n\nNeoadjuvant therapy for HER2 overexpressing breast cancer\n\nHER2 overexpression is a good predictive marker of HER2 targeted therapy, which means that HER2 therapy will be very effective in reducing the size of HER2 positive breast cancers. Therefore, the size of HER2 positive breast cancer can easily be reduced in patients who wish to have a breast-conserving operation, and potentially improve the outcome of patients if pCR can be achieved. pCR in HER2 overexpressing breast cancer after neoadjuvant therapy is associated with improved survival.\n\nA randomized phase II study (CHER-LOB) showed that the combination of lapatinib and trastuzumab is superior in achieving breast-conserving surgery or pCR in HER2 positive breast cancer patients, compared to either trastuzumab or lapatinib alone in combination with 12 weeks of paclitaxel followed by FEC chemotherapy. The rates of breast-conserving surgery were 66.7%, 57.9%, and 68.9% in trastuzumab alone (arm A), lapatinib alone (arm B) and combination arm (arm C), respectively. The pCR rates were 25% (90% CI, 13.1% to 36.9%) in arm A, 26.3% (90% CI, 14.5% to 38.1%) in arm B, and 46.7% (90% CI, 34.4% to 58.9%) in arm C (exploratory P = .019), showing improved efficacy in double targeting arm107. The TECHNO trial also evaluated pre-operative EC (epirubicin+cyclophosphamide) followed by TH (paclitaxel+trastuzumab) in HER2 overexpressing breast cancer. The DFS of patients who achieved pCR was 88% compared to patients without pCR 73% (p=0.01). pCR was the only significant prognostic factor for DFS (HR 2.5; 95% CI 1.2 to 5.1; p=0.013) from multivariate analysis. Patients who did not achieve pCR had an increased risk for relapse and death108. NSABP B-41 trial studied single agent lapatinib combined with ACT, in comparison with dual HER2 blockade with lapatinib and trastuzumab and ACT. pCR was achieved in 52.5% of the patients in the trastuzumab arm versus 53.2% in the lapatinib arm, compared to 62% in the treatment arm with combination of trastuzumab and lapatinib, thus showing a significant improvement by using double targeting therapy to achieve pCR109.\n\nPertuzumab is a recombinant humanized monoclonal antibody that targets the extracellular dimerization domain (sub-domain II) of HER2, as well as binding to the ligand binding site of HER3110. To date, there have been 2 neoadjuvant trials to test the role of trastuzumab – Neosphere and Tryphaena111,112. The NeoSphere trial studied combination of dual HER2 blockade – with ertuzumab, trastuzumab, and docetaxel givenery 3 weeks for a total of 4 cycles. Following surgery, all patients received 3 cycles of FEC IV every 3 weeks and trastuzumab was administered IV every 3 weeks to complete 1 year of therapy. The trial’s primary endpoint was a pCR rate defined as the absence of invasive cancer in the breast (ypT0/is). The addition of pertuzumab resulted in increased rate of pCR, 45.8% [95% CI 36.1–55.7] compared from 29.0% [95% CI 20.6–38.5]111. Based on improved pCR rate, the US FDA approved the use of pertuzumab in combination with trastuzumab and docetaxel for both metastatic and neoadjuvant setting of HER2 positive breast cancers113.\n\n\nNeoadjuvant therapy for triple negative breast cancer patients\n\nThe role of neoadjuvants in TNBC subtype cancers is somewhat mixed, and non-linear. From an analysis of 1118 patients who received neoadjuvant chemotherapy at the MD Anderson Cancer Center for stage I-III breast cancer during 1985 to 2004, 23% (total of 255 patients) had TNBC. TNBC patients had higher pCR rates compare to non-TNBC patients, but had rather decreased 3 year progression free survival, OS, and post-recurrence survival. For patients who achieved pCR, the outcomes were similar in both groups. When patients had TNBC, the recurrence and death rates were higher in the first 3 years, and once they had recurrence, the survival was significantly worse81.\n\nThe NSABP-18 trial results showed that the breast-conserving success rate was higher after neoadjuvant chemotherapy, when compared to same baseline characters but without neoadjuvant therapy. However, the trial did not result in disease specific mortality advantage for stage II tumors. The NSABP B-27 study had three arms with AC or AC and docetaxel pre-operatively, or AC followed by post-operative docetaxel. In this study, the patients who received AC and docetaxel for 4 cycles pre-operatively had higher pCR rate; however this did not result in OS or DFS benefit. The administration of docetaxel post-operatively improved DFS in patients who had a partial response to pre-operative AC. All adjuvant chemotherapy regimens are thought to be appropriate to use in neoadjuvant settings as well103.\n\nA meta-analysis observed the clinical outcome of TNBC and non-TNBC patients who received platinum-based adjuvant therapy compared to those who did not receive platinum-based therapy. The clinical complete response and pCR rate were both higher in TNBC when platinum-based chemotherapy was used114. Based on accumulated data in neoadjuvant settings, platinum agents will be likely incorporated into the standard of care treatment of TNBC in the near future.\n\nAs previously mentioned, the carriers of BRCA1 and 2 mutations are susceptive of DNA breakage due to defective DNA repair machinery. Therefore, poly ADP ribose polymerase (PARP) inhibition, which could be a rescue mechanism of DNA repair when BRCA proteins are not available, creates a ‘synthetic lethality’ when given with DNA damaging agents. The I-SPY2 trial concluded that veliparib and carboplatin combination treatment arm for TNBC patients is eligible to be moved to the phase III trial, given the 90% probability of superiority, when compared to standard the chemotherapy arm115. Based on this promising result, this combination arm could be considered as standard of care.\n\n\nSummary and conclusion\n\nBreast cancer treatment has achieved the biggest strides in the improvement of survival over the last few decades. Unfortunately, many women still experience recurrence of disease, or metastasis of primary tumor after early stage tumor has been treated. A better understanding of the underlying biology of the heterogenic nature of breast cancer has already enabled the development of targeted therapy and profiling tools to reduce the disease recurrence and mortality rate caused by breast cancers. However, there are still many questions to be answered, and patients to be saved. Cooperative efforts of both basic science of discovery and development of novel strategy to target individual tumors, enhanced understanding of tumor biology, faster adaptation of novel therapy among treating health professionals, as well as novel design of clinical trials will further improve our odds in the war of breast cancer, a disease that still causes the death of 425,000 women each year world-wide.", "appendix": "Author contributions\n\n\n\nElizabeth Lauren Millerb wrote a manuscript as a first author – based on partly written paragraphs that other 4 middle authors have contributed to.\n\nHee Jin Lee, Amriti Lulla, Prashanth Gokare, Liz Hernandez: these four authors contributed equally.\n\nBora Lim: Planned the structure of the paper, chose the articles to be reviewed and finalized the final manuscript.\n\nAll authors 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\nReferences\n\nIARC GLOBOCAN 2012: estimated cancer incidence, mortality and prevalence worldwide in 2012. 2012. Reference Source\n\nBilal E, Dutkowski J, Guinney J, et al.: Improving breast cancer survival analysis through competition-based multidimensional modeling. PLOS Comput Biol. 2013; 9(5): e1003047. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLehmann BD, Bauer JA, Chen X, et al.: Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011; 121(7): 2750–2767. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoduc KD, Cheang MC, Tyldesley S, et al.: Breast cancer subtypes and the risk of local and regional relapse. J Clin Oncol. 2010; 28(10): 1684–1691. PubMed Abstract \| Publisher Full Text\n\nvan de Vijver MJ, He YD, van't Veer LJ, et al.: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med. 2002; 347(25): 1999–2009. PubMed Abstract \| Publisher Full Text\n\nvan’t Veer LJ, Dai H, van de Vijver MJ, et al.: Gene expression profiling predicts clinical outcome of breast cancer. Nature. 2002; 415(6871): 530–536. PubMed Abstract \| Publisher Full Text\n\nCarlson RW, Hudis CA, Pritchard KI: Adjuvant endocrine therapy in hormone receptor-positive postmenopausal breast cancer: evolution of NCCN, ASCO, and St Gallen recommendations. J Natl Compr Canc Netw. 2006; 4(10): 971–979. PubMed Abstract\n\nStanford JL, Szklo M, Brinton LA: Estrogen receptors and breast cancer. Epidemiol Rev. 1986; 8: 42–59. PubMed Abstract\n\nGoldhirsch A, Wood WC, Coates AS, et al.: Strategies for subtypes--dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol. 2011; 22(8): 1736–1747. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBergh J, Jonsson PE, Glimelius B, et al.: A systematic overview of chemotherapy effects in breast cancer. Acta Oncol. 2001; 40(2–3): 253–281. PubMed Abstract \| Publisher Full Text\n\nBonadonna G, Moliterni A, Zambetti M, et al.: 30 years' follow up of randomised studies of adjuvant CMF in operable breast cancer: cohort study. BMJ. 2005; 330(7485): 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGroup EBCTC. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet. 2005; 365(9472): 1687–1717. PubMed Abstract \| Publisher Full Text\n\nPeto R, Davies C, Godwin J, et al.: Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term outcome among 100,000 women in 123 randomised trials. Lancet. 2012; 379(9814): 432–444. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJatoi I, Hilsenbeck SG, Clark GM, et al.: Significance of axillary lymph node metastasis in primary breast cancer. J Clin Oncol. 1999; 17(8): 2334–2340. PubMed Abstract\n\nBonadonna G, Brusamolino E, Valagussa P, et al.: Combination chemotherapy as an adjuvant treatment in operable breast cancer. N Engl J Med. 1976; 294(8): 405–410. PubMed Abstract \| Publisher Full Text\n\nPalmieri C, Jones A: The 2011 EBCTCG polychemotherapy overview. Lancet. 2012; 379(9814): 390–392. PubMed Abstract \| Publisher Full Text\n\nFisher B, Redmond C, Wickerham DL, et al.: Doxorubicin-containing regimens for the treatment of stage II breast cancer: The National Surgical Adjuvant Breast and Bowel Project experience. J Clin Oncol. 1989; 7(5): 572–582. PubMed Abstract\n\nFisher B, Brown AM, Dimitrov NV, et al.: Two months of doxorubicin-cyclophosphamide with and without interval reinduction therapy compared with 6 months of cyclophosphamide, methotrexate, and fluorouracil in positive-node breast cancer patients with tamoxifen-nonresponsive tumors: results from the National Surgical Adjuvant Breast and Bowel Project B-15. J Clin Oncol. 1990; 8(9): 1483–1496. PubMed Abstract\n\nFisher B, Anderson S, Tan-Chiu E, et al.: Tamoxifen and chemotherapy for axillary node-negative, estrogen receptor-negative breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-23. J Clin Oncol. 2001; 19(4): 931–942. PubMed Abstract\n\nDignam JJ, Huang L, Ries L, et al.: Estimating breast cancer-specific and other-cause mortality in clinical trial and population-based cancer registry cohorts. Cancer. 2009; 115(22): 5272–5283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuss HB, Berry DA, Cirrincione CT, et al.: Adjuvant chemotherapy in older women with early-stage breast cancer. N Engl J Med. 2009; 360(20): 2055–2065. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrancis P, Crown J, Di Leo A, et al.: Adjuvant chemotherapy with sequential or concurrent anthracycline and docetaxel: Breast International Group 02-98 randomized trial. J Natl Cancer Inst. 2008; 100(2): 121–133. PubMed Abstract \| Publisher Full Text\n\nMackey JR, Martin M, Pienkowski T, et al.: Adjuvant docetaxel, doxorubicin, and cyclophosphamide in node-positive breast cancer: 10-year follow-up of the phase 3 randomised BCIRG 001 trial. Lancet Oncol. 2013; 14(1): 72–80. PubMed Abstract \| Publisher Full Text\n\nHenderson IC, Berry DA, Demetri GD, et al.: Improved outcomes from adding sequential Paclitaxel but not from escalating Doxorubicin dose in an adjuvant chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol. 2003; 21(6): 976–983. PubMed Abstract \| Publisher Full Text\n\nMamounas EP, Bryant J, Lembersky B, et al.: Paclitaxel after doxorubicin plus cyclophosphamide as adjuvant chemotherapy for node-positive breast cancer: results from NSABP B-28. J Clin Oncol. 2005; 23(16): 3686–3696. PubMed Abstract \| Publisher Full Text\n\nCitron ML, Berry DA, Cirrincione C, et al.: Randomized trial of dose-dense versus conventionally scheduled and sequential versus concurrent combination chemotherapy as postoperative adjuvant treatment of node-positive primary breast cancer: first report of Intergroup Trial C9741/Cancer and Leukemia Group B Trial 9741. J Clin Oncol. 2003; 21(8): 1431–1439. PubMed Abstract \| Publisher Full Text\n\nBudd GT, Barlow WE, Moore HCF, et al.: First analysis of SWOG S0221: A phase III trial comparing chemotherapy schedules in high-risk early breast cancer. -- Budd et al. 29 (15): 1004 -- ASCO Meeting Abstracts. J Clin Oncol. 2011; 29(15_suppl): s1004. Reference Source\n\nMartin M, Ruiz A, Ruiz Borrego M, et al.: Fluorouracil, doxorubicin, and cyclophosphamide (FAC) versus FAC followed by weekly paclitaxel as adjuvant therapy for high-risk, node-negative breast cancer: results from the GEICAM/2003-02 study. J Clin Oncol. 2013; 31(20): 2593–2599. PubMed Abstract \| Publisher Full Text\n\nDutertre M, Smith CL: Molecular mechanisms of selective estrogen receptor modulator (SERM) action. J Pharmacol Exp Ther. 2000; 295(2): 431–437. PubMed Abstract\n\nTrunet PF, Vreeland F, Royce C, et al.: Clinical use of aromatase inhibitors in the treatment of advanced breast cancer. J Steroid Biochem Mol Biol. 1997; 61(3–6): 241–245. PubMed Abstract\n\nJordan VC: Tamoxifen: a most unlikely pioneering medicine. Nat Rev Drug Discov. 2003; 2(3): 205–213. PubMed Abstract \| Publisher Full Text\n\nWaters EA, Cronin KA, Graubard BI, et al.: Prevalence of tamoxifen use for breast cancer chemoprevention among U.S. women. Cancer Epidemiol Biomarkers Prev. 2010; 19(2): 443–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavies C, Pan H, Godwin J, et al.: Long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years after diagnosis of oestrogen receptor-positive breast cancer: ATLAS, a randomised trial. Lancet. 2013; 381(9869): 805–816. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHowell A, Cuzick J, Baum M, et al.: Results of the ATAC (Arimidex, Tamoxifen, Alone or in Combination) trial after completion of 5 years' adjuvant treatment for breast cancer. Lancet. 2005; 365(9453): 60–62. PubMed Abstract \| Publisher Full Text\n\nGoss PE, Ingle JN, Pater JL, et al.: Late extended adjuvant treatment with letrozole improves outcome in women with early-stage breast cancer who complete 5 years of tamoxifen. J Clin Oncol. 2008; 26(12): 1948–1955. PubMed Abstract \| Publisher Full Text\n\nMouridsen H, Giobbie-Hurder A, Goldhirsch A, et al.: Letrozole therapy alone or in sequence with tamoxifen in women with breast cancer. N Engl J Med. 2009; 361(8): 766–776. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowsett M, Cuzick J, Ingle J, et al.: Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol. 2010; 28(3): 509–518. PubMed Abstract \| Publisher Full Text\n\nChao C, Studts JL, Abell T, et al.: Adjuvant chemotherapy for breast cancer: how presentation of recurrence risk influences decision-making. J Clin Oncol. 2003; 21(23): 4299–4305. PubMed Abstract \| Publisher Full Text\n\nWelcome to Adjuvant! Online. In. 2014.\n\nRavdin PM, Siminoff LA, Davis GJ, et al.: Computer program to assist in making decisions about adjuvant therapy for women with early breast cancer. J Clin Oncol. 2001; 19(4): 980–991. PubMed Abstract\n\nFisher B, Dignam J, Tan-Chiu E, et al.: Prognosis and treatment of patients with breast tumors of one centimeter or less and negative axillary lymph nodes. J Natl Cancer Inst. 2001; 93(2): 112–20. PubMed Abstract \| Publisher Full Text\n\nDowsett M, Cuzick J, Wale C, et al.: Prediction of risk of distant recurrence using the 21-gene recurrence score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: a TransATAC study. J Clin Oncol. 2010; 28(11): 1829–1834. PubMed Abstract \| Publisher Full Text\n\nPaik S, Tang G, Shak S, et al.: Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor–positive breast cancer. J Clin Oncol. 2006; 24(23): 3726–34. PubMed Abstract \| Publisher Full Text\n\nInstitute NC. The TAILORx Breast Cancer Trial - National Cancer Institute. 2010. Reference Source\n\nAlbain KS, Barlow WE, Shak S, et al.: Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal women with node-positive, oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective analysis of a randomised trial. Lancet Oncol. 2010; 11(1): 55–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGonzalez-Angulo A, Barlow E, Gralow J, et al.: SWOG S1007: A phase III randomized clinical trial of standard adjuvant endocrine therapy with or without chemotherapy in patients with one to three positive nodes, hormone receptor (HR)-positive, and HER2-negative breast cancer with recurrence score (RS) of 25 or less. 2011 ASCO annual meeting: suppl: abstr TPS104. J Clin Oncol. 2013; 29. Reference Source\n\nBastien RR, Rodriguez-Lescure A, Ebbert MT, et al.: PAM50 breast cancer subtyping by RT-qPCR and concordance with standard clinical molecular markers. BMC Med Genomics. 2012; 5: 44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowsett M, Sestak I, Lopez-Knowles E, et al.: Comparison of PAM50 risk of recurrence score with oncotype DX and IHC4 for predicting risk of distant recurrence after endocrine therapy. J Clin Oncol. 2013; 31(22): 2783–90. PubMed Abstract \| Publisher Full Text\n\nAgendia. MammaPrint - MammaPrint - everything you need to know about MammaPrint. In. 2014.\n\nSchramek D, Leibbrandt A, Sigl V, et al.: Osteoclast differentiation factor RANKL controls development of progestin-driven mammary cancer. Nature. 2010; 468(7320): 98–102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobert E, Coleman H, Cameron D, et al.: Breast-cancer adjuvant therapy with zoledronic acid. N Engl J Med. 2014; 365(15): 1396–1405. PubMed Abstract \| Publisher Full Text\n\nPaterson AH, Anderson SJ, Lembersky BC, et al.: Oral clodronate for adjuvant treatment of operable breast cancer (National Surgical Adjuvant Breast and Bowel Project protocol B-34): a multicentre, placebo-controlled, randomised trial. Lancet Oncol. 2012; 13(7): 734–742. PubMed Abstract \| Publisher Full Text\n\nvon Minckwitz G, Untch M, Blohmer JU, et al.: Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol. 2012; 30(15): 1796–1804. PubMed Abstract \| Publisher Full Text\n\nvon Minckwitz GRM, Eidtmann H, Tesch H, et al.: The phase III NATAN study (GBC 36/ABCSG XX). In. 2014.\n\nGregory W, Marshall H, Bell R, et al.: Adjuvant zoledronic acid (ZOL) in postmenopausal women with breast cancer and those rendered postmenopausal: results of a meta-analysis. J Clin Oncol. 2012; 30. Reference Source\n\nYarden Y: Biology of HER2 and its importance in breast cancer. Oncology. 2001; 61(Suppl 2): 1–13. PubMed Abstract \| Publisher Full Text\n\nYarden Y, Shilo BZ: SnapShot: EGFR signaling pathway. Cell. 2007; 131(5): 1018. PubMed Abstract \| Publisher Full Text\n\nVogel C, Cobleigh MA, Tripathy D, et al.: First-line, single-agent Herceptin(R) (trastuzumab) in metastatic breast cancer. A preliminary report. Eur J Cancer. 2001; 37(Suppl 1): 25–29. PubMed Abstract \| Publisher Full Text\n\nBarok M, Isola J, Palyi-Krekk Z, et al.: Trastuzumab causes antibody-dependent cellular cytotoxicity-mediated growth inhibition of submacroscopic JIMT-1 breast cancer xenografts despite intrinsic drug resistance. Mol Cancer Ther. 2007; 6(7): 2065–2072. PubMed Abstract \| Publisher Full Text\n\nSlamon D, Eiermann W, Robert N, et al.: Adjuvant trastuzumab in HER2-positive breast cancer. N Engl J Med. 2011; 365(14): 1273–1283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoensuu H, Bono P, Kataja V, et al.: Fluorouracil, epirubicin, and cyclophosphamide with either docetaxel or vinorelbine, with or without trastuzumab, as adjuvant treatments of breast cancer: final results of the FinHer Trial. J Clin Oncol. 2009; 27(34): 5685–5692. PubMed Abstract \| Publisher Full Text\n\nGoldhirsch A, Gelber RD, Piccart-Gebhart MJ, et al.: 2 years versus 1 year of adjuvant trastuzumab for HER2-positive breast cancer (HERA): an open-label, randomised controlled trial. Lancet. 2013; 382(9897): 1021–1028. PubMed Abstract \| Publisher Full Text\n\nPerez EA, Suman VJ, Davidson NE, et al.: HER2 testing by local, central, and reference laboratories in specimens from the North Central Cancer Treatment Group N9831 intergroup adjuvant trial. J Clin Oncol. 2006; 24(19): 3032–3038. PubMed Abstract \| Publisher Full Text\n\nSpielmann M, Roche H, Delozier T, et al.: Trastuzumab for patients with axillary-node-positive breast cancer: results of the FNCLCC-PACS 04 trial. J Clin Oncol. 2009; 27(36): 6129–6134. PubMed Abstract \| Publisher Full Text\n\nHortobagyi GN, Perez EA: Integration of trastuzumab into adjuvant systemic therapy of breast cancer: ongoing and planned clinical trials. Semin Oncol. 2001; 28(5 Suppl 16): 41–46. PubMed Abstract \| Publisher Full Text\n\nPiccart-Gebhart MJ, Procter M, Leyland-Jones B, et al.: Trastuzumab after adjuvant chemotherapy in HER2–positive breast cancer. N Engl J Med. 2005; 353(16): 1659–1672. PubMed Abstract \| Publisher Full Text\n\nHolbro T, Beerli RR, Maurer F, et al.: The ErbB2/ErbB3 heterodimer functions as an oncogenic unit: ErbB2 requires ErbB3 to drive breast tumor cell proliferation. Proc Natl Acad Sci U S A. 2003; 100(15): 8933–8938. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRexer BN, Arteaga CL: Intrinsic and acquired resistance to HER2-targeted therapies in HER2 gene-amplified breast cancer: mechanisms and clinical implications. Crit Rev Oncog. 2012; 17(1): 1–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuw LY, O'Brien C, Pandita A, et al.: Acquired PIK3CA amplification causes resistance to selective phosphoinositide 3-kinase inhibitors in breast cancer. Oncogenesis. 2013; 2: e83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLoi S: Tumor-infiltrating lymphocytes, breast cancer subtypes and therapeutic efficacy. Oncoimmunology. 2013; 2(7): e24720. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClynes RA, Towers TL, Presta LG, et al.: Inhibitory Fc receptors modulate in vivo cytotoxicity against tumor targets. Nat Med. 2000; 6(4): 443–446. PubMed Abstract \| Publisher Full Text\n\nZhang D, Pal A, Bornmann W, et al.: Activity of lapatinib is independent of EGFR expression level in HER2–overexpressing breast cancer cells. Mol Cancer Ther. 2008; 7(7): 1846–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnston SR, Leary A: Lapatinib: a novel EGFR/HER2 tyrosine kinase inhibitor for cancer. Drugs Today (Barc). 2006; 42(7): 441–453. PubMed Abstract \| Publisher Full Text\n\nGoss PE, Smith IE, O'Shaughnessy J, et al.: Adjuvant lapatinib for women with early-stage HER2–positive breast cancer: a randomised, controlled, phase 3 trial. Lancet Oncol. 2013; 14(1): 88–96. PubMed Abstract \| Publisher Full Text\n\nMetzger-Filho O, Winer EP, Krop I: Pertuzumab: optimizing HER2 blockade. Clin Cancer Res. 2013; 19(20): 5552–5556. PubMed Abstract \| Publisher Full Text\n\nde Ruijter TC, Veeck J, de Hoon JP, et al.: Characteristics of triple-negative breast cancer. J Cancer Res Clin Oncol. 2011; 137(2): 183–192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHudis CA, Gianni L: Triple-negative breast cancer: an unmet medical need. Oncologist. 2011; 16(Suppl 1): 1–11. PubMed Abstract \| Publisher Full Text\n\nHudis CA, Gianni L: Triple-negative breast cancer: an unmet medical need. Oncologist. 2011; 16(Suppl 1): 1–11. PubMed Abstract \| Publisher Full Text\n\nClarke M, Coates AS, Darby SC, et al.: Adjuvant chemotherapy in oestrogen-receptor-poor breast cancer: patient-level meta-analysis of randomised trials. Lancet. 2008; 371(9606): 29–40. PubMed Abstract \| Publisher Full Text\n\nMartin M, Segui MA, Anton A, et al.: Adjuvant docetaxel for high-risk, node-negative breast cancer. N Engl J Med. 2010; 363(23): 2200–2210. PubMed Abstract \| Publisher Full Text\n\nLiedtke C, Mazouni C, Hess KR, et al.: Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer. J Clin Oncol. 2008; 26(8): 1275–1281. PubMed Abstract \| Publisher Full Text\n\nYoung SR, Pilarski RT, Donenberg T, et al.: The prevalence of BRCA1 mutations among young women with triple-negative breast cancer. BMC Cancer. 2009; 9: 86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKwon JS, Gutierrez-Barrera AM, Young D, et al.: Expanding the criteria for BRCA mutation testing in breast cancer survivors. J Clin Oncol. 2010; 28(27): 4214–4220. PubMed Abstract \| Publisher Full Text\n\nLinderholm B, Hellborg H, Johansson U, et al.: Significantly higher levels of vascular endothelial growth factor (VEGF) and shorter survival times for patients with primary operable triple-negative breast cancer. Ann Oncol. 2009; 20(10): 1639–1646. PubMed Abstract \| Publisher Full Text\n\nGreenberg S, Rugo HS: Triple-negative breast cancer: role of antiangiogenic agents. Cancer J. 2010; 16(1): 33–38. PubMed Abstract \| Publisher Full Text\n\nCameron D, Brown J, Dent R, et al.: Adjuvant bevacizumab-containing therapy in triple-negative breast cancer (BEATRICE): primary results of a randomised, phase 3 trial. Lancet Oncol. 2013; 14(10): 933–942. PubMed Abstract \| Publisher Full Text\n\nDeo SV, Bhutani M, Shukla NK, et al.: Randomized trial comparing neo-adjuvant versus adjuvant chemotherapy in operable locally advanced breast cancer (T4b N0-2 M0). J Surg Oncol. 2003; 84(4): 192–197. PubMed Abstract \| Publisher Full Text\n\nSweeting RS, Klauber-Demore N, Meyers MO, et al.: Young women with locally advanced breast cancer who achieve breast conservation after neoadjuvant chemotherapy have a low local recurrence rate. Am Surg. 2011; 77(7): 850–855. PubMed Abstract\n\nAkay CL, Meric-Bernstam F, Hunt KK, et al.: Evaluation of the MD Anderson Prognostic Index for local-regional recurrence after breast conserving therapy in patients receiving neoadjuvant chemotherapy. Ann Surg Oncol. 2012; 19(3): 901–907. PubMed Abstract \| Publisher Full Text\n\nMatuschek C, Bolke E, Roth SL, et al.: Long-term outcome after neoadjuvant radiochemotherapy in locally advanced noninflammatory breast cancer and predictive factors for a pathologic complete remission: results of a multivariate analysis. Strahlenther Onkol. 2012; 188(9): 777–781. PubMed Abstract \| Publisher Full Text\n\nKaufmann M, von Minckwitz G, Mamounas EP, et al.: Recommendations from an international consensus conference on the current status and future of neoadjuvant systemic therapy in primary breast cancer. Ann Surg Oncol. 2012; 19(5): 1508–1516. PubMed Abstract \| Publisher Full Text\n\nvan der Hage JA, van de Velde CJ, Julien JP, et al.: Preoperative chemotherapy in primary operable breast cancer: results from the European Organization for Research and Treatment of Cancer trial 10902. J Clin Oncol. 2001; 19(22): 4224–4237. PubMed Abstract\n\nMauri D, Pavlidis N, Ioannidis JP: Neoadjuvant versus adjuvant systemic treatment in breast cancer: a meta-analysis. J Natl Cancer Inst. 2005; 97(3): 188–194. PubMed Abstract \| Publisher Full Text\n\nKrainick-Strobel UE, Lichtenegger W, Wallwiener D, et al.: Neoadjuvant letrozole in postmenopausal estrogen and/or progesterone receptor positive breast cancer: a phase IIb/III trial to investigate optimal duration of preoperative endocrine therapy. BMC Cancer. 2008; 8: 62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThomas E, Holmes FA, Smith TL, et al.: The use of alternate, non-cross-resistant adjuvant chemotherapy on the basis of pathologic response to a neoadjuvant doxorubicin-based regimen in women with operable breast cancer: long-term results from a prospective randomized trial. J Clin Oncol. 2004; 22(12): 2294–2302. PubMed Abstract \| Publisher Full Text\n\nCameron DA, Anderson ED, Levack P, et al.: Primary systemic therapy for operable breast cancer--10-year survival data after chemotherapy and hormone therapy. Br J Cancer. 1997; 76(8): 1099–1105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCameron DA, Anderson ED, Levack P, et al.: Primary systemic therapy for operable breast cancer--10-year survival data after chemotherapy and hormone therapy. Br J Cancer. 1997; 76(8): 1099–1105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSemiglazov V, Semiglazov V, Ivanov V, et al.: The relative efficacy of neoadjuvant endocrine therapy vs chemotherapy in postmenopausal women with ER- positive breast cancer. J Clin Oncol. 2004; 22(14S): 519. Reference Source\n\nSmith IE, Dowsett M, Ebbs SR, et al.: Neoadjuvant treatment of postmenopausal breast cancer with anastrozole, tamoxifen, or both in combination: the Immediate Preoperative Anastrozole, Tamoxifen, or Combined with Tamoxifen (IMPACT) multicenter double-blind randomized trial. J Clin Oncol. 2005; 23(22): 5108–5116. PubMed Abstract \| Publisher Full Text\n\nBaselga J, Semiglazov V, van Dam P, et al.: Phase II randomized study of neoadjuvant everolimus plus letrozole compared with placebo plus letrozole in patients with estrogen receptor-positive breast cancer. J Clin Oncol. 2009; 27(16): 2630–2637. PubMed Abstract \| Publisher Full Text\n\nSataloff DM, Mason BA, Prestipino AJ, et al.: Pathologic response to induction chemotherapy in locally advanced carcinoma of the breast: a determinant of outcome. J Am Coll Surg. 1995; 180(3): 297–306. PubMed Abstract\n\nFisher B, Bryant J, Wolmark N, et al.: Effect of preoperative chemotherapy on the outcome of women with operable breast cancer. J Clin Oncol. 1998; 16(8): 2672–2685. PubMed Abstract\n\nBear H: Primary chemotherapy for operable breast cancer: the NSABP experience. Breast Cancer Research. 2005; 7(Suppl 1): S17. Publisher Full Text\n\nBear HD, Anderson S, Brown A, et al.: The effect on tumor response of adding sequential preoperative docetaxel to preoperative doxorubicin and cyclophosphamide: preliminary results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol. 2003; 21(22): 4165–4174. PubMed Abstract \| Publisher Full Text\n\nLoibl S, von Minckwitz G, Raab G, et al.: Surgical procedures after neoadjuvant chemotherapy in operable breast cancer: results of the GEPARDUO trial. Ann Surg Oncol. 2006; 13(11): 1434–1442. PubMed Abstract \| Publisher Full Text\n\nSteger EK, Hausmaninger H, Gnant M, et al.: 6 vs. 3 cycles of epirubicin/docetaxel + G-CSF in operable breast cancer: results of ABCSG-14. Steger, et al. 22 (14 Supplement): 553 -- ASCO Meeting Abstracts. J Clin Oncol. 2004; 22(suppl 553). Reference Source\n\nGuarneri V, Frassoldati A, Bottini A, et al.: Preoperative chemotherapy plus trastuzumab, lapatinib, or both in human epidermal growth factor receptor 2-positive operable breast cancer: results of the randomized phase II CHER-LOB study. J Clin Oncol. 2012; 30(16): 1989–1995. PubMed Abstract \| Publisher Full Text\n\nUntch M, Fasching PA, Konecny GE, et al.: Pathologic complete response after neoadjuvant chemotherapy plus trastuzumab predicts favorable survival in human epidermal growth factor receptor 2-overexpressing breast cancer: results from the TECHNO trial of the AGO and GBG study groups. J Clin Oncol. 2011; 29(25): 3351–3357. PubMed Abstract \| Publisher Full Text\n\nRobidoux A, Tang G, Rastogi P, et al.: Lapatinib as a component of neoadjuvant therapy for HER2-positive operable breast cancer (NSABP protocol B-41): an open-label, randomised phase 3 trial. Lancet Oncol. 2013; 14(12): 1183–1192. PubMed Abstract \| Publisher Full Text\n\nKeating GM: Pertuzumab: in the first-line treatment of HER2-positive metastatic breast cancer. Drugs 2012; 72(3): 353–360. PubMed Abstract \| Publisher Full Text\n\nGianni L, Pienkowski T, Im YH, et al.: Efficacy and safety of neoadjuvant pertuzumab and trastuzumab in women with locally advanced, inflammatory, or early HER2-positive breast cancer (NeoSphere): a randomised multicentre, open-label, phase 2 trial. Lancet Oncol. 2012; 13(1): 25–32. PubMed Abstract \| Publisher Full Text\n\nSchneeweiss A, Chia S, Hickish T, et al.: Pertuzumab plus trastuzumab in combination with standard neoadjuvant anthracycline-containing and anthracycline-free chemotherapy regimens in patients with HER2-positive early breast cancer: a randomized phase II cardiac safety study (TRYPHAENA). Ann Oncol. 2013; 24(9): 2278–2284. PubMed Abstract \| Publisher Full Text\n\nYao S: FDA approves Perjeta for neoadjuvant breast cancer treatment. 2013. Reference Source\n\nLiu M, Mo QG, Wei CY, et al.: Platinum-based chemotherapy in triple-negative breast cancer: A meta-analysis. Oncol Lett. 2013; 5(3): 983–991. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRugo HS, Olopade O, DeMichele A, et al.: Veliparib/carboplatin plus standard neoadjuvant therapy for high-risk breast cancer: First efficacy results from the I-SPY2 trial. Canc Res. 2013; 72(Suppl). Reference Source" }	[ { "id": "5867", "date": "28 Aug 2014", "name": "Daniel Vorobiof", "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 overview on the current therapy of breast cancer, divided according to different specific stages and patients and tumor characteristics.The abstract as well as the rest of the text provide a detailed explanation of the different clinical breast cancer stages and sub-stages and its corresponding medical treatments. The authors have reviewed the extensive relevant published literature and, when possible, referenced with the latest publications in the specific subsets.This publication will be of value to medical students, fellows in the areas of clinical medicine and for specialists with an interest in breast cancer. It is well written and each section appropriately referenced.", "responses": [] }, { "id": "6636", "date": "04 Nov 2014", "name": "Ahmad Awada", "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 very concise paper given the key messages from a wide and evolving topic. The authors provide important information and summary for clinicians.I’m supporting the indexation of this paper.SuggestionsTitle: I propose adding: key messages from adjuvant and neoadjuvant therapyAbstract: One minor comment: last sentence of the abstract: … developing early treatments for different…", "responses": [] } ]	1	https://f1000research.com/articles/3-198
https://f1000research.com/articles/2-256/v1	25 Nov 13	{ "type": "Review", "title": "A review of the biologic and pharmacological role of docosapentaenoic acid", "authors": [ "Puya G Yazdi" ], "abstract": "Fish oil contains a complex mixture of omega-3 fatty acids, of which eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), and docosahexaenoic acid (DHA) are the three predominant forms. There has been a plethora of previous research on the effects and associations of fish oil supplementation with various clinical manifestations. While the majority of this work was previously done on EPA and DHA, emerging research has begun to elucidate the specific role that DPA plays in these physiological processes and its differences with the other omega-3 fatty acids. The purpose of this review is to focus on the new studies undertaken with DPA. This review summarizes the biochemical mechanisms involved in the biosynthesis and metabolism of DPA before focusing on its effects in cardiovascular disease, immune function, and psychiatric and cognitive health. The limited studies point toward a positive role that DPA supplementation can play in these processes and that is separate and distinct from traditional supplementation with DHA and EPA.", "keywords": [ "As a result of anecdotal reports of their low incidence of coronary heart disease", "Bang and Dyerburg began to study the Greenland Eskimo (Inuits) population in the late 1960s. Their pioneering findings confirmed the anecdotal evidence", "Inuits had lower incidences of myocardial infarction", "better lipid profiles", "reduced platelet activity", "and lower incidence of immune and inflammatory diseases compared with western controls1–3. These findings were attributed to the Inuit diet", "and specifically to the large quantities of seal and whale meat consumed by the Inuits. Eventually it was deduced that marine n-3 fatty acids found in the seal and whale meat was the main protective agent against cardiovascular heart disease4. These findings", "along with separate research that demonstrated mammalian brain grey matter was also rich in n-3 fatty acids", "became an impetus for much scientific and clinical research into the potential health benefits of n-3 fatty acids5", "6. Over the ensuing years", "n-3 fatty acids have been found to hold great therapeutic promise in a myriad of conditions including", "but not limited to", "neural function", "diabetes mellitus", "cardiovascular health", "cancer", "lipid regulation", "and as a anti-inflammatory agents7–12." ], "content": "Introduction\n\nAs a result of anecdotal reports of their low incidence of coronary heart disease, Bang and Dyerburg began to study the Greenland Eskimo (Inuits) population in the late 1960s. Their pioneering findings confirmed the anecdotal evidence; Inuits had lower incidences of myocardial infarction, better lipid profiles, reduced platelet activity, and lower incidence of immune and inflammatory diseases compared with western controls1–3. These findings were attributed to the Inuit diet, and specifically to the large quantities of seal and whale meat consumed by the Inuits. Eventually it was deduced that marine n-3 fatty acids found in the seal and whale meat was the main protective agent against cardiovascular heart disease4. These findings, along with separate research that demonstrated mammalian brain grey matter was also rich in n-3 fatty acids, became an impetus for much scientific and clinical research into the potential health benefits of n-3 fatty acids5,6. Over the ensuing years, n-3 fatty acids have been found to hold great therapeutic promise in a myriad of conditions including, but not limited to, neural function, diabetes mellitus, cardiovascular health, cancer, lipid regulation, and as a anti-inflammatory agents7–12.\n\nThe majority of the studies conducted on n-3 fatty acids were conducted on fish oils that contain a mixture of the three n-3 fatty acids: eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA) and docosahexaenoic acid, (DHA). While much attention has been focused on DHA and EPA, the literature on DPA remains brief. In recent years, research on DPA has increased. Much of the increased interest in DPA came about from two observations. First, the seal meat so readily consumed by the Inuits contains high concentrations of DPA; and second, human mother’s milk also contains high concentrations of DPA13. There already exists extensive literature and reviews on the beneficial effects of EPA and DHA which the reader is encouraged to read14–18. The following review will aim to address some basic biological insights behind DPA and some of its potential beneficial effects. Additionally, for a more thorough review of the biochemical and molecular pathways related to DPA, the reader is encouraged to read another recent DPA review, cited in the biosynthesis section below. This review will focus on the clinical associations involved with DPA and it will do this by focusing on DPA specific studies in addition to studies conducted on fish oils or mixtures of omega-3 n-fatty acids that contained DPA. Studies in which the role of DPA was studied specifically will be noted as well studies that were conducted on animal models or in vitro.\n\n\nBiosynthesis and metabolism of essential polyunsaturated fatty acids\n\nEPA, DHA, and DPA are the three major polyunsaturated fatty acids formed by a series of desaturation and elongation enzymes from Alpha-linolenic acid (ALA)19. Fatty acid elongase-2 (FAE-2) and FAE-5 are the two enzymes responsible for the direct conversion of EPA into DPA by direct chain elongation20–22. The conversion of DPA into DHA is more circuitous involving chain elongation followed by desaturation in the cytoplasm before being moved to the peroxisome to be chain shortened by b-oxidation to form DHA23,24. It is worth noting that in living cells this process can go in reverse yielding DPA from EPA, a process that most likely involves peroxisomal acyl-coA and b-oxidation25 For a more thorough review of the biochemistry behind DPA the reader is encouraged to read the previously aforementioned DPA review25. Figure 1 summarizes the major pathways involved in n-3 fatty acid production.\n\nSummary of the biochemical enzymes and intermediates involved in n-fatty acid production in cells. The blue underneath the lipid bilayer represents a peroxisome with steps occurring outside in the cytoplasm.\n\n\nClinical associations of DPA\n\nNumerous studies have reported possible anti-cancer effects of n-3 fatty acids, particularly for breast, colon, and prostate cancer26–28. Omega-3 fatty acids reduced prostate tumor growth, slowed pathological progression, and increased survival in mice29. Among n-3 fatty acids, high levels of DHA, which is the most abundant n-3 polyunsaturated fat in erythrocyte membranes, were associated with a reduced risk of breast cancer30. Additionally, a 2007 systematic review of n-3 fatty acids and cachexia found evidence that oral n-3 fatty acid supplementation was beneficial as adjuvant cancer therapy by improving appetite, weight, quality of life, and retaining muscle mass31,32. Additionally, a recent study looking at the anti-tumorigenic effects of n-3 fatty acids on colorectal cancer found anti-proliferative and pro-apoptotic effects for all 3 fatty acids, with DPA demonstrating the strongest effects in both in vitro and in vivo models of colorectal cancer33.\n\nNot surprisingly, some of the main initial work was done on the effects of n-3 fatty acid and DPA supplementation in heart disease and specifically myocardial infarction. As early as three months and onwards, treatment with 1gram per day of n-3 fatty acids resulted in a statistically significant reduction of the occurrence of death, sudden cardiac death, and cardiovascular death34,35. Furthermore, it has also been demonstrated that n-3 fatty acids have a statistically significant antihypertensive effect by lowering systolic blood pressure by 3.5–5.5 mmHg36. For a more thorough review of all the potential benefits of n-3 fatty acids in general on the cardiovascular system, the reader is advised to read a recent review18.\n\nWhile much of this previous work has shown that omega-3 fatty acids can play a protective role in maintaining a healthy cardiovascular system, recent studies have demonstrated a positive correlation between DPA itself and cardiovascular disease prevention in humans37–39. A prospective population study of 1871 subjects in Eastern Finland demonstrated that subjects with plasma blood concentrations of DPA plus DHA in the 20th percentile had a decreased relative risk of acute coronary events of 44%, compared to those subjects in bottom 20th percentile. The decreased relative risk rose to 67% so long as the top 20th percentile group had mercury levels below or equal to 2 micrograms/g compared to those in the bottom 20th percentile who had mercury levels above 2 micrograms/g38. This study was followed up by a Japanese study, published five years later that further corroborated these findings by showing a significant association between DPA supplementation and reduced cardiovascular disease37. A nested-case control study of 6438 adults with seven years of follow up also showed that DPA reduced the odds ratio of heart disease40. More importantly, this same finding was further confirmed by another nested-case control study of 32,826 subjects with 6 years of follow up that also demonstrated that this association was independent of the other n-3 fatty acids41. Interestingly, a cross-sectional study of 26 subjects and 24 matched controls revealed that decreased DPA concentrations in the cell membrane of erythrocytes (3.0+/-0.19% for subjects vs. 3.9+/-0.12% for controls, P<0.001) were statistically significantly associated with increased heart disease42.\n\nRecent studies have further developed our understanding of which pathophysiologic mechanisms DPA supplementation can target in cardiovascular disease. First, a prospective cohort study which directly measured circulating levels of n-3 fatty acids in the blood of 2735 US adults without existing heart disease who were enrolled in the Cardiovascular Health Study from 1992 to 2006 found that total n-3 fatty concentration was associated with lower incidence of congestive heart failure with DPA, conferring as much as a 40% reduction43. A second study looked at the impact of n-3 fatty acid on coronary plaque instability by use of conventional and integrated backscatter intravascular ultrasound approaches44. Their results demonstrated that low serum concentrations of DPA were significantly associated with lipid-rich plaques, suggesting that decreased DPA levels can contribute to increased incidence of plaque formation leading to acute coronary syndrome and myocardial infarction44. Besides heart disease, stroke prevention and arterial blockage also comprise a significant part of cardiovascular health. Plasma levels of DPA are inversely related to blocked arteries and DPA the only n-3 fatty acid that has been shown to significantly reduce blocked arteries regardless of lifetime smoking45. A cross-sectional study with carotid ultrasonography revealed that carotid artery wall thickness decreased with DPA intake46.\n\nAdditionally, in vitro work has started to isolate the biochemical mechanisms by which DPA could help combat cardiovascular disease. For instance, platelet aggregation is an early event in the development of thrombosis and is initiated by thromboxane A247. Recently, an in vitro study conducted in rabbit platelets showed that EPA, DPA and DHA inhibited collagen- or arachidonic acid stimulated platelet aggregation in a dose dependent manner, with DPA being the most potent inhibitor and possibly en times more powerful than EPA in inhibiting platelet aggregation48. A further study conducted on human whole blood corroborated these earlier findings49. Endothelial cell (EC) migration and proliferation are important processes in the control of the wound-healing response of blood vessels50. DPA has been shown to be a potent stimulator of EC migration51. Furthermore, DPA pretreatment of bovine aortic endothelial (BAE) cells inhibits their migrating activity due to vascular endothelial growth factor (VEGF) stimulation9. Additionally, that same pretreatment suppresses tube formation demonstrating that DPA is a potent inhibitor of angiogenesis, a physiological mechanism that contributes to tumor growth, inflammation, and microangiopathy9.\n\nNumerous studies have demonstrated that EPA and DHA can lower triglycerides (TG) and cholesterol levels in the plasma and liver, by increasing β-oxidation activity in the mitochondria and peroxisomes in hepatic cells, and suppressing TG synthesis in the liver52–54. Similar work is now showing that DPA also possesses lipid metabolism improving effects similar to EPA and DHA in improving cholesterol and TG levels. A recent in vivo study demonstrated that DPA can reduce non-HDL cholesterol by 50%55. Finally, DPA has been shown to have a positive role in reducing the expression of inflammatory genes (inflammation in the walls of blood vessels is thought to play a role in the development of atherosclerotic plaques leading to cardiovascular disease) thereby improving cardiovascular health56.\n\nThe role of n-3 fatty acid supplementation in immune function is just beginning to emerge as a new frontier in fatty acid research. In a study regarding fish oil supplementation in infants, the authors found that fish oil supplementation could lead to quicker immune maturation without a concomitant reduction in immune activation57. Additionally, owing to the anti-inflammatory properties of n-3 fatty acids and the pro-inflammatory properties of n-6 fatty acids found in most western diets, many researchers currently believe that fish oil supplementation can aid many chronic inflammatory conditions by decreasing the n-6 to n-3 fatty acid ratio58. As one example, patients suffering from rheumatoid arthritis report reduced pain symptoms when taking n-3 fatty acids in conjunction with NSAIDS compared to those only taking NSAIDS59. Finally, enzymatically oxygenated derivatives (oxylipins) of DPA have been shown to be potent anti-inflammatory compounds in animal models60.\n\nThere is now emerging evidence that n-3 fatty acids can play a role in psychiatric disorders due to the observation that schizophrenic patients demonstrate reduced levels of both n-3 and n-6 fatty acids61. Treating high-risk children with a dietary supplement of n-3 fatty acids demonstrated a statistically significant decrease in progression to schizophrenia61. Additionally, it has also been shown that patients with schizophrenia have decreased levels of DPA in erythrocytes62. Finally, a meta-analysis based on 10 clinical trials, found that n-3 fatty acids significantly improved depression in patients with both unipolar and bipolar disorder63.\n\nNeurological and cognitive decline as a result of aging or disease is now emerging as a new avenue of further research in n-3 fatty acid supplementation. First, in a rat model of Alzheimer's disease, EPA supplementation revealed a statistically significant efficacy in countering memory impairment64. This study spurred interest in studying the possible benefit of DPA supplementation in cognitive decline due to aging. Specifically, during the aging process, there is a loss of synaptic function which leads to deficits in spatial learning tasks and reduced ability of rats to sustain long term potentiation65. By supplementing the normal diets of rats with DPA it was found that DPA possesses neuro-restorative effects in the hippocampus by decreasing microglial activation and oxidative stress, the two major biochemical mechanisms involved in cognitive decline due to synaptic function loss66. These authors concluded that DPA supplementation might play a significant role in neuro-protection against age-related cognitive decline by attenuating the age-related decline in spatial learning and long-term potentiation.\n\n\nConclusions\n\nWhile much work has been done on the potential therapeutic benefits of n-3 fatty acids, the majority of that work has been done on EPA and DHA. Initially, the Inuit population and their diet caught the attention of the medical community because of their much lower incidence of cardiovascular disease, which was attributed to their consumption of n-3 fatty acids. What has been forgotten in the ensuing years was that their seal meat also had high concentrations of DPA in addition to the more familiar EPA and DHA counterparts. Additionally, levels of DPA in human breast milk are high, and DPA levels in adult human blood are similar to EPA13. These findings are a part of a growing body of work on DPA with promising results. Additionally, this research has also begun to elucidate important differences between DPA and EPA and DHA. Specifically, DPA inhibits platelet aggregation more efficiently than EPA or DHA, DPA stimulates endothelial cell migration much more efficiently than EPA, and finally DPA is incorporated into human plasma and red blood cell lipids faster than EPA and hence may act as a reservoir of the major n-3 fatty acids in humans9,51,67. Studies looking into DPA and its clinical associations are beginning to demonstrate that lack of DPA in diets and blood circulation may serve as an independent predictor and marker for various health conditions. The existing evidence points to DPA as showing potential as a nutritional and therapeutic supplement. While much work still needs to be addressed on the possible benefits of DPA consumption in human health and disease, the limited data available seems to indicate that DPA can have additional health benefits in conjunction with the more common n-3 fatty acids.", "appendix": "Competing interests\n\n\n\nPGY is a consultant for Cyvex Nutrition Inc, which is a distributor of nutritional supplements based on DPA and other n-3 fatty acids.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe author would like to thank Jared F Taylor for helpful discussions on the manuscript in addition to formatting help and Luciano Herrera for help on the colorful illustration.\n\n\nReferences\n\nBang HO, Dyerberg J, Nielson AB: Plasma Lipid and Lipoprotein Pattern in Greenlandic West-Coast Eskimos. Lancet. 1971; 1(7710): 1143–1145. PubMed Abstract \| Publisher Full Text\n\nDyerberg J, Bang HO, Aagaard O: Alpha-Linolenic Acid and Eicosapentaenoic Acid. Lancet. 1980; 1(8161): 199. PubMed Abstract \| Publisher Full Text\n\nKromann N, Green A: Epidemiological-Studies in the Upernavik District, Greenland - Incidence of Some Chronic Diseases 1950–1974. Acta Med Scand. 1980; 208(5): 401–406. PubMed Abstract \| Publisher Full Text\n\nDyerberg J, Bang HO, Stoffersen E, et al.: Eicosapentanoic Acid and Prevention of Thrombosis and Atherosclerosis. Lancet. 1978; 2(8081): 117–119. PubMed Abstract \| Publisher Full Text\n\nCrawford MA, Sinclair AJ: Nutritional influences in the evolution of mammalian brain. In: lipids, malnutrition & the developing brain. Ciba Found Symp. 1971; 267–292. PubMed Abstract\n\nCrawford MA, Casperd NM, Sinclair AJ: The Long-Chain Metabolites of Linoleic and Linolenic Acids in Liver and Brain in Herbivores and Carnivores. Comp Biochem Physiol B. 1976; 54(3): 395–401. PubMed Abstract \| Publisher Full Text\n\nArita M, Yoshida M, Hong S, et al.: Resolvin E1, an endogenous lipid mediator derived from omega-3 eicosapentaenoic acid, protects against 2,4,6-trinitrobenzene sulfonic acid-induced colitis. Proc Natl Acad Sci U S A. 2005; 102(21): 7671–7676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAktas H, Halperin JA: Translational regulation of gene expression by omega-3 fatty acids. J Nutr. 2004; 134(9): 2487S–2491S. PubMed Abstract\n\nTsuji M, Murota SI, Morita I: Docosapentaenoic acid (22:5, n-3) suppressed tube-forming activity in endothelial cells induced by vascular endothelial growth factor. Prostaglandins Leukot Essent Fatty Acids. 2003; 68(5): 337–342. PubMed Abstract \| Publisher Full Text\n\nKitajka K, Puskas LG, Zvara A, et al.: The role of n-3 polyunsaturated fatty acids in brain: modulation of rat brain gene expression by dietary n-3 fatty acids. Proc Natl Acad Sci U S A. 2002; 99(5): 2619–2624. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFujikawa M, Yamazaki K, Hamazaki T, et al.: Effect of eicosapentaenoic acid ethyl ester on albuminuria in streptozotocin-induced diabetic rats. J Nutr Sci Vitaminol (Tokyo). 1994; 40(1): 49–61. PubMed Abstract\n\nShimizu H, Ohtani K, Tanaka Y, et al.: Long-term effect of eicosapentaenoic acid ethyl (EPA-E) on albuminuria of non-insulin dependent diabetic patients. Diabetes Res Clin Pract. 1995; 28(1): 35–40. PubMed Abstract \| Publisher Full Text\n\nKoletzko B, Mrotzek M, Bremer HJ: Fatty acid composition of mature human milk in Germany. Am J Clin Nutr. 1988; 47(6): 954–959. PubMed Abstract\n\nDavidson MH: Omega-3 fatty acids: new insights into the pharmacology and biology of docosahexaenoic acid, docosapentaenoic acid, and eicosapentaenoic acid. Curr opin lipidol. 2013; 24(6): 467–474. PubMed Abstract \| Publisher Full Text\n\nKaur G, Sinclair AJ, Cameron-Smith D, et al.: Docosapentaenoic acid (22: 5n-3) down-regulates the expression of genes involved in fat synthesis in liver cells. Prostaglandins Leukot Essent Fatty Acids. 2011; 85(3–4): 155–161. PubMed Abstract \| Publisher Full Text\n\nLovegrove JA, Griffin BA: The acute and long-term effects of dietary fatty acids on vascular function in health and disease. Curr Opin Clin Nutr Metab Care. 2013; 16(2): 162–167. PubMed Abstract \| Publisher Full Text\n\nNicholson T, Khademi H, Moghadasian MH: The role of marine n-3 fatty acids in improving cardiovascular health: a review. Food Funct. 2013; 4(3): 357–365. PubMed Abstract \| Publisher Full Text\n\nDe Caterina R: n-3 fatty acids in cardiovascular disease. N Engl J Med. 2011; 364(25): 2439–2450. PubMed Abstract \| Publisher Full Text\n\nMohrhauer H, Holman RT: Tracer Experiments to Assess Metabolic Conversions of Polyunsaturated Fatty Acids. J Am Oil Chem Soc. 1965; 42: 639–643. PubMed Abstract \| Publisher Full Text\n\nHorton JD, Shah NA, Warrington JA, et al.: Combined analysis of oligonucleotide microarray data from transgenic and knockout mice identifies direct SREBP target genes. Proc Natl Acad Sci U S A. 2003; 100(21): 12027–12032. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Y, Botolin D, Christian B, et al.: Tissue-specific, nutritional, and developmental regulation of rat fatty acid elongases. J Lipid Res. 2005; 46(4): 706–715. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoss A, Reinhart M, Sankarappa S, et al.: The metabolism of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,13,16,19-docosahexaenoic acid in rat liver is independent of a 4-desaturase. J Biol Chem. 1991; 266(30): 19995–20000. PubMed Abstract\n\nChristensen E, Woldseth B, Hagve TA, et al.: Peroxisomal beta-oxidation of polyunsaturated long chain fatty acids in human fibroblasts. The polyunsaturated and the saturated long chain fatty acids are retroconverted by the same acyl-CoA oxidase. Scand J Clin Lab Invest Suppl. 1993; 215: 61–74. PubMed Abstract\n\nReddy JK, Hashimoto T: Peroxisomal beta-oxidation and peroxisome proliferator-activated receptor alpha: an adaptive metabolic system. Annu Rev Nutr. 2001; 21: 193–230. PubMed Abstract \| Publisher Full Text\n\nKaur G, Cameron-Smith D, Garg M, et al.: Docosapentaenoic acid (22: 5n-3): a review of its biological effects. Prog Lipid Res. 2011; 50(1): 28–34. PubMed Abstract \| Publisher Full Text\n\nAugustsson K, Michaud DS, Rimm EB, et al.: A prospective study of intake of fish and marine fatty acids and prostate cancer. Cancer Epidemiol Biomarkers Prev. 2003; 12(1): 64–67. PubMed Abstract\n\nde Deckere EA: Possible beneficial effect of fish and fish n-3 polyunsaturated fatty acids in breast and colorectal cancer. Eur J Cancer Prev. 1999; 8(3): 213–221. PubMed Abstract\n\nCaygill CP, Hill MJ: Fish, n-3 fatty acids and human colorectal and breast cancer mortality. Eur J Cancer Prev. 1995; 4: 329–332. PubMed Abstract\n\nBerquin IM, Min Y, Wu R, et al.: Modulation of prostate cancer genetic risk by omega-3 and omega-6 fatty acids. J Clin Invest. 2007; 117(7): 1866–1875. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPala V, Krogh V, Muti P, et al.: Erythrocyte membrane fatty acids and subsequent breast cancer: a prospective Italian study. J Natl Cancer Inst. 2001; 93(14): 1088–1095. PubMed Abstract\n\nColomer R, Moreno-Nogueira JM, García-Luna PP, et al.: N-3 fatty acids, cancer and cachexia: a systematic review of the literature. Br J Nutr. 2007; 97(5): 823–831. PubMed Abstract \| Publisher Full Text\n\nRyan AM, Reynolds JV, Healy L, et al.: Enteral nutrition enriched with eicosapentaenoic acid (EPA) preserves lean body mass following esophageal cancer surgery: results of a double-blinded randomized controlled trial. Ann Surg. 2009; 249(3): 355–363. PubMed Abstract \| Publisher Full Text\n\nMorin C, Rousseau E, Fortin S: Anti-proliferative effects of a new docosapentaenoic acid monoacylglyceride in colorectal carcinoma cells. Prostaglandins Leukot Essent Fatty Acids. 2013; 89(4): 203–213. PubMed Abstract \| Publisher Full Text\n\nDietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico. Lancet. 1999; 354(9177): 447–455. PubMed Abstract \| Publisher Full Text\n\nMarchioli R, Barzi F, Bomba E, et al.: Early protection against sudden death by n-3 polyunsaturated fatty acids after myocardial infarction: time-course analysis of the results of the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico (GISSI)-Prevenzione. Circulation. 2002; 105(16): 1897–1903. PubMed Abstract \| Publisher Full Text\n\nAppel LJ, Miller ER 3rd, Seidler AJ, et al.: Does supplementation of diet with 'fish oil' reduce blood pressure? A meta-analysis of controlled clinical trials. Arch Intern Med. 1993; 153(12): 1429–1438. PubMed Abstract \| Publisher Full Text\n\nOda E, Hatada K, Katoh K, et al.: A case-control pilot study on n-3 polyunsaturated fatty acid as a negative risk factor for myocardial infarction. Int Heart J. 2005; 46(4): 583–591. PubMed Abstract \| Publisher Full Text\n\nRissanen T, Voutilainen S, Nyyssonen K, et al.: Fish oil-derived fatty acids, docosahexaenoic acid and docosapentaenoic acid, and the risk of acute coronary events: the Kuopio ischaemic heart disease risk factor study. Circulation. 2000; 102(22): 2677–2679. PubMed Abstract \| Publisher Full Text\n\nOda E, Hatada K, Kimura J, et al.: Relationships between serum unsaturated fatty acids and coronary risk factors: negative relations between nervonic acid and obesity-related risk factors. Int Heart J. 2005; 46(6): 975–985. PubMed Abstract \| Publisher Full Text\n\nSimon JA, Hodgkins ML, Browner WS, et al.: Serum fatty acids and the risk of coronary heart disease. Am J Epidemiol. 1995; 142(5): 469–476. PubMed Abstract\n\nSun Q, Ma J, Campos H, et al.: Blood concentrations of individual long-chain n-3 fatty acids and risk of nonfatal myocardial infarction. Am J Clin Nutr. 2008; 88(1): 216–223. PubMed Abstract\n\nPaganelli F, Maixent JM, Duran MJ, et al.: Altered erythrocyte n-3 fatty acids in Mediterranean patients with coronary artery disease. Int J Cardiol. 2001; 78(1): 27–32. PubMed Abstract \| Publisher Full Text\n\nMozaffarian D, Lemaitre RN, King IB, et al.: Circulating long-chain omega-3 fatty acids and incidence of congestive heart failure in older adults: the cardiovascular health study: a cohort study. Ann Intern Med. 2011; 155(3): 160–170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmano T, Matsubara T, Uetani T, et al.: Impact of omega-3 polyunsaturated fatty acids on coronary plaque instability: an integrated backscatter intravascular ultrasound study. Atherosclerosis. 2011; 218(1): 110–116. PubMed Abstract \| Publisher Full Text\n\nLeng GC, Horrobin DF, Fowkes FG, et al.: Plasma essential fatty acids, cigarette smoking, and dietary antioxidants in peripheral arterial disease. A population-based case-control study. Arterioscler Thromb. 1994; 14(3): 471–478. PubMed Abstract \| Publisher Full Text\n\nHino A, Adachi H, Toyomasu K, et al.: Very long chain N-3 fatty acids intake and carotid atherosclerosis: an epidemiological study evaluated by ultrasonography. Atherosclerosis. 2004; 176: 145–149. PubMed Abstract \| Publisher Full Text\n\nJackson SP: The growing complexity of platelet aggregation. Blood. 2007; 109(12): 5087–5095. PubMed Abstract \| Publisher Full Text\n\nAkiba S, Murata T, Kitatani K, et al.: Involvement of lipoxygenase pathway in docosapentaenoic acid-induced inhibition of platelet aggregation. Biol Pharm Bull. 2000; 23(11): 1293–1297. PubMed Abstract\n\nPhang M, Garg ML, Sinclair AJ: Inhibition of platelet aggregation by omega-3 polyunsaturated fatty acids is gender specific-Redefining platelet response to fish oils. Prostaglandins Leukot Essent Fatty Acids. 2009; 81(1): 35–40. PubMed Abstract \| Publisher Full Text\n\nLamalice L, Le Boeuf F, Huot J: Endothelial cell migration during angiogenesis. Circulation research. 2007; 100: 782–794.Publisher Full Text\n\nKanayasu-Toyoda T, Morita I, Murota S: Docosapentaenoic acid (22:5, n-3), an elongation metabolite of eicosapentaenoic acid (20:5, n-3), is a potent stimulator of endothelial cell migration on pretreatment in vitro. Prostaglandins Leukot Essent Fatty Acids. 1996; 54(5): 319–325. PubMed Abstract \| Publisher Full Text\n\nGottlicher M, Widmark E, Li Q, et al.: Fatty acids activate a chimera of the clofibric acid-activated receptor and the glucocorticoid receptor. Proc Natl Acad Sci U S A. 1992; 89(10): 4653–4657. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHorton JD, Goldstein JL, Brown MS: SREBPs: transcriptional mediators of lipid homeostasis. Cold Spring Harb Symp Quant Biol. 2002; 67: 491–498. PubMed Abstract \| Publisher Full Text\n\nPawar A, Jump DB: Unsaturated fatty acid regulation of peroxisome proliferator-activated receptor alpha activity in rat primary hepatocytes. J Biol Chem. 2003; 278(38): 35931–35939. PubMed Abstract \| Publisher Full Text\n\nChen JN, Jiang Y, Liang Y, et al.: DPA n-3, DPA n-6 and DHA improve lipoprotein profiles and aortic function in hamsters fed a high cholesterol diet. Atherosclerosis. 2012; 221(2): 397–404. PubMed Abstract \| Publisher Full Text\n\nKishida E, Tajiri M, Masuzawa Y: Docosahexaenoic acid enrichment can reduce L929 cell necrosis induced by tumor necrosis factor. Biochim Biophys Acta. 2006; 1761(4): 454–462. PubMed Abstract \| Publisher Full Text\n\nDamsgaard CT, Lauritzen L, Kjaer TM, et al.: Fish oil supplementation modulates immune function in healthy infants. J Nutr. 2007; 137(4): 1031–1036. PubMed Abstract\n\nWall R, Ross RP, Fitzgerald GF, et al.: Fatty acids from fish: the anti-inflammatory potential of long-chain omega-3 fatty acids. Nutr Rev. 2010; 68(5): 280–289. PubMed Abstract \| Publisher Full Text\n\nRuggiero C, Lattanzio F, Lauretani F, et al.: Omega-3 polyunsaturated fatty acids and immune-mediated diseases: inflammatory bowel disease and rheumatoid arthritis. Curr Pharm Des. 2009; 15(36): 4135–4148. PubMed Abstract \| Publisher Full Text\n\nDangi B, Obeng M, Nauroth JM, et al.: Biogenic synthesis, purification, and chemical characterization of anti-inflammatory resolvins derived from docosapentaenoic acid (DPAn-6). J Biol Chem. 2009; 284: 14744–14759. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmminger GP, Schäfer MR, Papageorgiou K, et al.: Long-chain omega-3 fatty acids for indicated prevention of psychotic disorders: a randomized, placebo-controlled trial. Arch Gen Psychiatry. 2010; 67(2): 146–154. PubMed Abstract \| Publisher Full Text\n\nAssies J, Lieverse R, Vreken P, et al.: Significantly reduced docosahexaenoic and docosapentaenoic acid concentrations in erythrocyte membranes from schizophrenic patients compared with a carefully matched control group. Biol Psychiatry. 2001; 49(6): 510–522. PubMed Abstract \| Publisher Full Text\n\nLin PY, Su KP: A meta-analytic review of double-blind, placebo-controlled trials of antidepressant efficacy of omega-3 fatty acids. J Clin Psychiatry. 2007; 68(7): 1056–1061. PubMed Abstract \| Publisher Full Text\n\nTaepavarapruk P, Song C: Reductions of acetylcholine release and nerve growth factor expression are correlated with memory impairment induced by interleukin-1beta administrations: effects of omega-3 fatty acid EPA treatment. J Neurochem. 2010; 112(4): 1054–1064. PubMed Abstract \| Publisher Full Text\n\nRosenzweig ES, Barnes CA: Impact of aging on hippocampal function: plasticity, network dynamics, and cognition. Prog Neurobiol. 2003; 69(3): 143–179. PubMed Abstract \| Publisher Full Text\n\nKelly L, Grehan B, Chiesa AD, et al.: The polyunsaturated fatty acids, EPA and DPA exert a protective effect in the hippocampus of the aged rat. Neurobiol Aging. 2011; 32(12): 2318.e1–2315. PubMed Abstract \| Publisher Full Text\n\nMiller E, Kaur G, Larsen A, et al.: A short-term n-3 DPA supplementation study in humans. Eur J Nutr. 2013; 52(3): 895–904. PubMed Abstract \| Publisher Full Text" }	[ { "id": "3468", "date": "17 Feb 2014", "name": "Philip Calder", "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 remains significant interest in the health impact of omega-3 fatty acids from oily fish and fish oil type supplements on human health. The physiologic and health benefits are usually ascribed to EPA and DHA. However a third fatty acid, DPA, is present and this has been relatively little researched. Some new work on DPA has been performed which has suggested that it has some unique functional properties. These are reviewed elsewhere, but new updates, views and insights are always to be welcomed. This article aims to provide a current review of DPA and, I think, to distinguish it from EPA and DHA. To be frank, I do not think it achieves the latter aim as well as it might (see comments below).Specific comments:Title. If the \"biologic\" is used then this should be countered by \"pharmacologic\". Title. Should clarify that this is about DPA n-3 not DPA n-6. Abstract. We are told in line 1 that fish oil contains a complex mixture of EPA, DPA and DHA. Therefore it cannot be correct to state on line 5 that ‘It is not correct to state that work with fish oil \"was previously done on EPA and DHA\". Perhaps the author means that the work has focussed on EPA and DHA as the active components?\n\nThe last sentence of the abstract makes the same error. We are told first (line 1) that fish oil contains EPA, DPA and DHA. Thus it cannot be correct to state that traditional supplementation is \"with DHA and EPA\" - it is with all three marine omega-3 fatty acids. Perhaps the author means that DPA has \"separate and distinct properties compared with EPA and DHA\"? Page 2, column 1, 5 and 2 lines from the bottom. ‘b’ in b-oxidation should be shown as the Greek symbol beta. Page 2, column 1, 3 lines from bottom. I think \"DPA from EPA\" should read \"DPA from DHA\" since the former is not \"reverse\" metabolism and would not involve beta-oxidation. Page 2, column 2, line 3. What is meant by \"Clinical associations\"? Does the author mean \"Clinical effects\"? The author is setting out to establish that DPA has independent functional features. However in many parts the evidence for this is based upon rehearsing the effects of fish oil supplements - these data cannot be used to say DPA has effects, since the effects may be due to any one of EPA, DPA and DHA or indeed to all three. Too often effects are due to \"n-3 fatty acids\" and it is then concluded that DPA is functional. I would much rather that the author is explicit in picking out and describing those studies where DPA is shown to have a function. A really poor example is in the first paragraph about cardiovascular disease where the effects seen in the GISSI trial are described. It is NOT acknowledged that these can have nothing to do with DPA since the preparation used in that study was a pharmaceutical preparation composed almost entirely of ethyl esters of EPA and DHA. To say otherwise is to mislead readers. Page 4, column 1, line 31. Should \"en\" read \"ten\"? Page 4, column 1, 4 lines from bottom. Please add \"in hamsters\" after \"50%\".", "responses": [] }, { "id": "4785", "date": "19 May 2014", "name": "Bruce Bistrian", "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 general this article does make a good point that DPA may well be an important omega 3 fatty acid and provides references from in vitro studies, animal studies, and some human epidemiologic studies that DPA may have some unique, beneficial properties. However many of the studies quoted for support in cardiovascular disease confuse association with causation, and this should be noted. Moreover there is still some controversy about the extent of the effectiveness of very long chain omega 3 fatty acids in the improvement of cardiovascular health according to different health authorities. That said the article does provide a useful review that should promote more research into this potentially important fatty acid. As an additional comment there are some misstatements which are presumably typographic on the first page. The retroconversion is DPA to EPA rather than the reverse. Also there is only one elongase necessary to convert EPA to DHA.", "responses": [] } ]	1	https://f1000research.com/articles/2-256
https://f1000research.com/articles/3-197/v1	19 Aug 14	{ "type": "Research Article", "title": "Causes of elective surgery cancellation and theatre throughput efficiency in an Australian urology unit", "authors": [ "Andrew Keller", "Akbar Ashrafi", "Ahmad Ali", "Akbar Ashrafi", "Ahmad Ali" ], "abstract": "Objective:To evaluate our unit’s theatre throughput efficiency, to identify where inefficiencies existed and consequently where the greatest improvement might be made.To identify the causes of day of surgery cancellations and how they might be avoided. Patients and Methods:A prospective audit of theatre utilisation was undertaken over a 6 month period between 05/02//2013 and 02/08/2013 at Ipswich General Hospital, QLD, Australia.Times collected were: time of patient arrival in anaesthetic bay, start time of operative procedure, end time of operative procedure, and time of patient leaving theatre.The causative factors for any delays or day of surgery cancellations were identified and recorded where possible. Results:In the six month period 26,850 sessional minutes were available for elective operating over 100 operating sessions.304 elective cases were performed, split between 21 major and 283 minor proceduresThe sessions ran overtime a cumulative 2114 minutes.Total non-operative minutes totalled 13,209 (50.3% of all available time), split between late starts 499 minutes (1.8%), early list finishes 1894 minutes (7.05%), changeover time 1869 minutes (6.9%) and anaesthetic time, 8974 minutes (33.4%)Actual operating time only compromised 50.7% of all available elective operating session time (13,614 minutes)Theatre utilisation was 91.8%.51 procedures were cancelled on the day of surgery during the audit period, representing 14.3% of all scheduled procedures.The most common reason for cancellation was lack of surgical fitness, followed by inadequate operative time. Conclusion: A significant proportion of all elective operative time was consumed by non-operative minutes.Inefficiencies existed in turnover of patients as well as over as well as underbooking of patients on elective lists.An excessive number of cases were cancelled on the day of surgery, wasting valuable operative time.A multi-parametric approach must be taken to improve operation list utilisation.", "keywords": [ "Theatre efficiency is increasingly coming under the spotlight as elective waiting lists continue to increase1. Delays or interruptions during operating lists are associated with dissatisfaction for health care providers and patients alike2. Theatre lists account for a significant proportion of a hospital’s revenue and an even larger fraction of its total expenses3–6. As operating theatre budgets are already stretched", "increased case throughput must come from improved theatre efficiency rather than from more operating sessions." ], "content": "Introduction\n\nTheatre efficiency is increasingly coming under the spotlight as elective waiting lists continue to increase1. Delays or interruptions during operating lists are associated with dissatisfaction for health care providers and patients alike2. Theatre lists account for a significant proportion of a hospital’s revenue and an even larger fraction of its total expenses3–6. As operating theatre budgets are already stretched, increased case throughput must come from improved theatre efficiency rather than from more operating sessions.\n\nEfficient use of theatre sessions relies on prompt start times, an appropriately booked theatre case-mix, efficient patient turnover, and finishing on time to reduce overtime costings7,8. Accurate scheduling of elective theatre cases to maximise operating efficiency is extremely complex, as the time required for identical procedures can vary dramatically.\n\nThe most cost-efficient method to increase theatre case throughput is by decreasing idle theatre time9,10.\n\n\nMethod\n\nWe sought to evaluate our unit’s theatre throughput efficiency so we might identify areas where the most time was wasted during operating sessions, and consequently where the most significant improvements might be made.\n\nTo do this we undertook a prospective audit of all elective theatre operating in the Urology unit at Ipswich General Hospital (IGH), a regional secondary referral hospital, over a six month period, between 05/02//2013 and 02/08/2013. The theatre complex at IGH consists of 6 operating suites with a 3 bay arrangement, with each suite having an anaesthetic bay and scrub room in addition to the operating room itself.\n\nTimes were extracted from the Operating Room Management Information System (ORMIS) theatre management software (CSC). Versions 5 and subsequently 7 were used, as the software was updated during the audit period. Times extracted from ORMIS were: time of arrival in anaesthetic bay, start time of operative procedure, end time of operative procedure, and time of patient leaving theatre (Table 1). Times were entered into ORMIS by theatre nursing staff as per standard practice.\n\nWhere possible the reasons for delays were identified and recorded by both the nursing staff entering the reasons into ORMIS, and by direct recording by a surgical team observer.\n\nThe scrub nurses, theatre assistants, and members of the anaesthetic team were not informed that the study was being conducted, so as not to influence their performance. The surgical team were undertaking the audit, and they were never blinded.\n\n\nResults\n\nIn the six month audit period 304 elective cases were performed, split between 21 major (Table 2) and 283 minor procedures (Table 3). Total available operative minutes were 26,850 distributed over 100 elective operating lists. Ordinarily each week consisted of one 3.5 hour, one 8.5 hour and two 3 hour sessions.\n\nStart time efficiency or late starts (LS) (Table 4), measured from the time the patient entered the anaesthetic bay, was acceptable at 499 minutes or 1.8% of all available list time (Table 5). Significant LS (over 15 minutes late) occurred on only 8 lists. Over all operating lists the mean LS was 5 minutes with a median of 0 minutes, and a range of 0–148 minutes. The total LS time was skewed significantly by a single episode where 6 nursing staff were absent with illness simultaneously, delaying the start of the list by 148 minutes. Only a small number of cases were delayed by the late arrival of anaesthetic or surgical team members. Delay in patients arriving from the day surgery unit or wards were more common but still infrequent.\n\nJust over 50% of all available sessional time was used for operating during the audit period.\n\n1894 minutes were wasted with early finishing (EF) (Table 4) of lists representing 7.05% of all available time (Table 5). Significant EF, considered as lists finishing over 15 minutes early, affected 24 lists and totalled 1830 minutes, with a mean of 76.25 minutes, a median of 45 and a range of 18–480 minutes.\n\nUnder booking of theatre lists accounted for a significant proportion of all EF, however, the lion’s share of early list finishes were caused by day of surgery cancellations. 27% of all EF (843 minutes) were accounted for by cancellation of just 3 cases (Table 6). Two radical retro-pubic prostatectomies (RRP) were cancelled due to patients changing their mind on the day of surgery and instead opting for external beam radiation therapy, with a cumulative loss of 363 minutes of scheduled sessional time (ST) (Table 4). The cancellation of a radical cystectomy, which was the only booked case on an all-day operating list, accounted for 480 lost minutes.\n\nCancellation reasons are classified as per Argo et al’s cancellation codes 11 .\n\nIn total, 51 procedures were cancelled on the day of surgery during the audit period, representing 14.3% of all scheduled procedures. The reasons for case cancellation were grouped into 5 categories, and 28 potential cancellation reasons, as per Argo et al’s audit of elective operating in the US Veteran’s Health Administration11 (Table 7). The most common reason for case cancellation was lack of fitness for surgery (W4), with inadequate operative time the second most common (M7) (Table 8).\n\nSource: Adapted from: Argo JL, Vick CC, Graham LA, Itani KM, Bishop MJ, Hawn MT. Elective surgical case cancellation in the Veterans Health Administration system: identifying areas for improvement. Am J Surg 2009;198:600–6.\n\nCancellation reasons are classified as per Argo et al’s cancellation codes11.\n\n2114 minutes were recorded of theatre overtime (OT) (Table 4), measured from the time patients left the operating room. This represented 7.9% overtime over the scheduled ST during the audit period (Table 5). Operative OT, measured from completion of the last operative procedure accounted for 1404 minutes (66.4%), with anaesthetic overtime accounting for the remaining 710 minutes (33.6%). Significant OT affected 37 operating lists, with a mean of 54.59 and median of 37 minutes and a range of 16–105 minutes. The causes of OT during the audit were multifactorial. Any unforeseen delays during the operative list, such as a slow patient changeover, difficult induction of anaesthesia, late start of the operative list, or unexpectedly prolonged operative time all contributed to total overtime.\n\nOne of the major contributors that was identified was major cases being booked onto half day operating lists. Two lists each week were of only 180 minutes duration, and on 4 out of 5 occasions when a major case was booked onto such a list, the session ran significantly overtime.\n\nTotal patient changeover time (CT) (Table 4), which was defined as the time the patient left the operating room until the subsequent patient on the operating list entered the anaesthetic bay was acceptable at 1869 minutes, representing 6.9% of all available operative list time (Table 5). The mean CT was 8.16 minutes, the median 5 minutes with a range of 0–132 minutes. A significant delay in CT, defined as those taking over 15 minutes, occurred on 31 occasions (13.4% of all changeovers). Late patient arrival or non-arrival at the day of surgery admissions was responsible for a significant proportion of all CT. The next patient was already in the anaesthetic bay before completion of the prior case on 60 occasions (26% of all changeovers), significantly reducing total CT.\n\nAnaesthetic time (AT) (Table 4) which consisted of: patient time spent in the anaesthetic bay, anaesthetic induction time and the time for the patient to leave theatre after the end of the procedure, totalled 9657 minutes. After removing the AT spent in overtime (710 minutes), AT consumed 33.4% of all available operating list time. Mean anaesthetic time over the audit was 29.52 minutes, with a median of 21.5 and a range of 2–64 minutes.\n\nOf all available time, 15018 minutes were spent operating (PT) (Table 4). After excluding the PT occurring after the scheduled end of the operating list (1404 minutes), this meant that only 50.7% of all available sessional time (ST) was spent operating (Table 5).\n\nTheatre utilisation over the entire audit period was 91.8%, however, this number was significantly skewed by the large amount of both OT and early start minutes (ES). ES, measured from entry of the patient into the anaesthetic bay prior to the scheduled start of the operating list, totalled 967 minutes. When these minutes, in addition to the OT (2114 minutes), are subtracted effective theatre utilisation falls to 80.3%.\n\nThere were several limitations to our study. While we attempted to blind the theatre assistants, anaesthetic and nursing staff from the ongoing audit, several members of each team became aware of the audit throughout its course. This could have influenced their efforts throughout the audit period. As all times for the study extracted from ORMIS were entered by nursing staff, it is possible that bias could have affected the accuracy of the times if the nurses entering the data were aware of the audit. The surgical team was never blinded to the audit, and this might have influenced the operative urgency of the surgeons involved and their punctuality.\n\nIn regards to our case mix, our relatively small proportion of major cases compared to a tertiary referral urology service would certainly increase the ratio of non-operative to operative time compared to an operative case mix with more major cases.\n\nAnother factor that influenced our throughput was addition of emergency cases to our elective lists. Ipswich General Hospital has one emergency list daily, which preferentially performs all emergency cases unless elective sessions finish early, or if the patient is medically unstable. During the audit period 4 emergency cases were added to the end of our elective list: 1 drainage of a scrotal abscess and 3 ureteric stents. Total overtime generated by these additional cases totalled 106 minutes. If these additional cases had not been performed, an additional 114 early finishing minutes would have been recorded.\n\n\nDiscussion\n\nOur audit has highlighted the complexity of maximising operative efficiency. Optimisation of theatre throughput efficiency starts with careful booking of the operating list. Currently operating lists are booked by ex-clinical staff with a best-guess approach, which while practical, often fails to take into account the myriad variables of the case and staffing mix. In various centres mathematical theories previously applied to the manufacturing industry have been successfully trialled to facilitate more efficient booking of theatre time, however, such methods require significant expertise and staff retraining and are not currently viable options at our institution12.\n\nDay of surgery cancellations affected 14.3% of all scheduled cases during our audit period. Whilst this number is not dissimilar to Argo et al’s analysis of urological cancellations in the Veterans Health Administration (14%), it still represents a significant amount of wasted ST11. Whilst some of these cancellations were unavoidable, such as staff illness, the majority of the 6 cancellations secondary to patient factors (P1, P6, P8, P10) could potentially have been avoided by a phone call to the patient a day or two prior to the operative date, allowing adequate time for replacement cases to be found. The 11 cancellations due to workup and administration factors (W1, W2, W4B, W5, F6) could foreseeably have been avoided by better communication between members of the surgical, anaesthetic and nursing teams. If foreseen early on the day of surgery, these otherwise wasted operative minutes could have been filled with elective patients called in at short notice. Other units have established a “fillbuster” list for just such events9. Whilst there was a significant proportion of our operative time wasted with day of surgery cancellations, we do acknowledge that implementing such a waitlist for elective surgery patients is logistically challenging in a smaller hospital such as ours, and is impractical for many of our patients.\n\nDuring the audit, CT consumed 6.9% of all available sessional time, and whilst the mean CT was acceptable at 8 minutes, on 31 occasions a significant delay of over 15 minutes occurred. Harders et al. successfully reduced their turnover time by 37% by supplying pagers for all theatre assistants which were triggered 5 minutes prior to completion of the case in addition to standardisation of equipment2. Whilst not an issue in our centre during the course of the audit, Weinbroum et al’s audit has shown that up to 10% of all available operating time is wasted awaiting Post-Anaesthetic Care Unit (PACU) space13. Wasted operative time costs an estimated $US 10–30 per minute or at least $US 600 per hour13–15. As the hourly cost of maintaining a single patient in PACU costs $US 110, increased throughput and profitability could both be achieved by increasing the staffing and by physical enlargement of the PACU13,16–18.\n\nThe greatest gains in terms of surgical throughput efficiency have been seen with implementation of parallel processing. Parallel processing involves preparing patients for theatre concurrently as the prior patient’s procedure is completed, contrasting to the traditional approach of serially processing patients. This approach allows for reduction of both AT and CT. Parallel processing allows intravenous and arterial lines to be inserted, and spinal or even general anaesthesia to be achieved in pre-procedure rooms5,19. Simultaneous processing often requires additional anaesthetic staffing and the consequent increased costs associated. However, the increased cost of implementing parallel processing can be offset by increased throughput and consequent financial gain6.\n\nParallel processing is most advantageous in operating lists where multiple, small cases are going to be performed, in cases with a consistent operative duration, and on full day operating lists20. These incremental time savings over preceding cases enable the performance of additional cases19,21–25. Whilst parallel processing only facilitates added cases on high turnover lists, it has been shown to reduce overtime costings in theatres where fewer, longer cases are performed, however, in this circumstance this might not offset the added costs of increased staffing levels22,26. Sandberg et al. have shown that parallel processing, when used across an entire theatre complex, is cost neutral23. More selective use of parallel processing, such as solely for high turnover lists, would yield the greatest benefit financially without affecting case throughput23.\n\n\nConclusion\n\nOur study quantified how much of each theatre session was occupied by non-operative processes in a unit with multiple short-duration operative procedures. Based on the findings of the study, we have implemented process changes to increase our own theatre efficiency. We have recently adopted a high-throughput theatre list one day a month, where multiple small simple cases are booked on an all-day operating list. Additional anaesthetic and nursing staff are rostered on these days to facilitate parallel processing of patients and remove the need for lunch breaks. If these prove to be successful and cost-efficient we would look to increase their frequency.\n\nWhilst we are unable to implement more complex booking algorithms at this stage due to financial constraints, we are still striving to improve our booking efficiency. Efforts are being made to more accurately document predicted operative times on booking forms. Uncomplicated patients are now having blood and urine taken when booked at outpatients. They are then being seen immediately by pre-operative nurses. This enables suitable patients to be prepared for surgery at short notice in the event of cancellations. Booking staff are now making efforts to contact patients several days prior to their surgery and in the event of problems, to liaise with the surgical team, to limit the number of day of surgery cancellations. Assessment of our unit’s theatre throughput is ongoing.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data files (Excel) for the audit of elective theatre operating in the Urology unit, Ipswich General Hospital, QLD, Australia, 10.5256/f1000research.4824.d3276627", "appendix": "Author contributions\n\n\n\nAndrew Keller - design of study, data gathering and writing. Akbar Ashrafi - design of study, data gathering and proofing. Ahmad Ali - Concept of study, design of study and proofing.\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\nDepartment of Health and Aging. The State of Our Public Hospitals June 2009 Report. Department of Health and Aging, Australian Government, Sydney. [Updated June 2009 Cited 26 Sept 2013] Available from URL. Reference Source\n\nHarders M, Malangoni MA, Weight S, et al.: Improving operating room efficiency through process redesign. Surgery. 2006; 140(4): 509–14. PubMed Abstract \| Publisher Full Text\n\nMacario A: What does one minute of operating room time cost? J Clin Anesth. 2010; 22(4): 233–6. PubMed Abstract \| Publisher Full Text\n\nHealthcare financial management association. Integration in a reform environment: strategies for success. [Updated June 2010 Cited 15 Sep 2012.] Available from URL: Reference Source.\n\nCima R, Brown M, Hebl R, et al.: Use of lean and six sigma methodology to improve operating room efficiency in a high-volume tertiary-care academic medical center. J Am Coll Surg. 2011; 213(1): 83–92. PubMed Abstract \| Publisher Full Text\n\nMarjamaa R, Vakkuri A, Kirvela O: Operating room management: why how and by whom? Acta Anaesthesiol Scand. 2008; 52(2): 596–600. PubMed Abstract \| Publisher Full Text\n\nBent S, Sherrier M, Peters M, et al.: Analyzing first-case starts utilising process engineering techniques. Can J Surg. 2010; 53: 167–70.\n\nTesti A, Tanfani E, Torre G: A three-phase approach for operating theatre schedules. Health Care Manag Sci. 2007; 10(2): 163–172. PubMed Abstract \| Publisher Full Text\n\nLehtonen JM, Kujala J, Kouri J, et al.: Cardiac surgery productivity and throughput improvements. Int J Health Care Qual Assur. 2007; 20(1): 40–52. PubMed Abstract \| Publisher Full Text\n\nBeattie C: Successful strategies for improving operating room efficiency at academic institutions. Anesth Analg. 1999; 88(4): 963–4. PubMed Abstract \| Publisher Full Text\n\nArgo JL, Vick CC, Graham LA, et al.: Elective surgical case cancellation in the Veterans Health Administration system: identifying areas for improvement. Am J Surg. 2009; 198(5): 600–606. PubMed Abstract \| Publisher Full Text\n\nHarrison S, Nugud O, Benziger H: Operating theatre management: do we lack a mathematician’s perspective? ANZ J Surg. 2013; 83(1–2): 5–6. PubMed Abstract \| Publisher Full Text\n\nWeinbroum AA, Ekstein P, Ezri T: Efficiency of the operating room suite. Am J Surg. 2003; 185(3): 244–50. PubMed Abstract \| Publisher Full Text\n\nMurray W, Schneider A: Using simulators for education and training in anesthesiology. ASA Newsletter. 1997; 61. Reference Source\n\nFarnworth LR, Lemay DE, Wooldridge T, et al.: A comparison of operative times in arthroscopic ACL reconstruction between orthopaedic faculty and residents: the financial impact of orthopaedic surgical training in the operating room. Iowa Orthop J. 2001; 21: 31–5. PubMed Abstract \| Free Full Text\n\nWaddle JP, Evers AS, Piccirillo JF: Postanesthesia care unit length of stay: quantifying and assessing dependent factors. Anesth Analg. 1998; 87(3): 628–33. PubMed Abstract \| Publisher Full Text\n\nLeslie J: A new technique for selective cost-effective PONV protocols utilizing a systems thinking analysis and resolution (STARTM) computer program for modeling and comparing patient care protocols. Anesthesiology. 1995; 83: A47.\n\nDelaney CL, David N, Tamblyn P: Audit of the utilization of time in an orthopaedic trauma theatre. ANZ J Surg. 2010; 80(4): 217–222. PubMed Abstract \| Publisher Full Text\n\nSokolovic E, Biro P, Wyss P, et al.: Impact of the reduction of anaesthesia turnover time on operating room efficiency. Eur J Anaesthesiol. 2002; 19(8): 560–3. PubMed Abstract \| Publisher Full Text\n\nFriedman DM, Sokal SM, Chang Y, et al.: Increased operating room efficiency through parallel processing. Ann Surg. 2006; 243(1): 10–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStahl JE, Sandberg WS, Daily B, et al.: Reorganizing patient care and workflow in the operating room: a cost-effectiveness study. Surgery. 2006; 139(6): 717–728. PubMed Abstract \| Publisher Full Text\n\nCendán JC, Good M: Interdisciplinary work flow assessment and redesign decreases operating room turnover time and allows for additional caseload. Arch Surg. 2006; 141(1): 65–9. PubMed Abstract \| Publisher Full Text\n\nSandberg WS, Daily B, Egan M, et al.: Deliberate perioperative systems design improves operating room throughput. Anesthesiology. 2005; 103(2): 406–18. PubMed Abstract \| Publisher Full Text\n\nTorkki PM, Marjamaa RA, Torkki MI, et al.: Use of anesthesia induction rooms can increase the number of urgent orthopedic cases completed within 7 hours. Anesthesiology. 2005; 103(2): 401–5. PubMed Abstract \| Publisher Full Text\n\nHanss R, Buttgereit B, Tonner PH, et al.: Overlapping induction of anesthesia: an analysis of benefits and costs. Anesthesiology. 2005; 103(2): 391–400. PubMed Abstract \| Publisher Full Text\n\nDexter F, Coffin S, Tinker JH: Decrease in anesthesia-controlled time cannot permit one additional surgical operation to be reliably scheduled during the workday. Anesth Analg. 1995; 81(6): 1263–8. PubMed Abstract \| Publisher Full Text\n\nKeller AT, Ashrafi A, Ali A: Raw data files for the audit of elective theatre operating in the Urology unit, Ipswich General Hospital, QLD, Australia. F1000Res. 2014. Data Source" }	[ { "id": "5860", "date": "28 Aug 2014", "name": "Levent Turkeri", "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 prospective study over a 6 month period to identify the causative factors for delays or cancellations including time of patient arrival in anaesthetic bay, start time of operative procedure, end time of operative procedure, and time of patient leaving theatre.In terms of non-operative time, anaesthetic process comprised the largest part with 8974 minutes (33.4%). Changeover time was approximately 7%. The most common reason for cancellation was lack of surgical fitness, followed by inadequate operative time.This study sheds light for possible reasons of inefficient OR time usage in a certain institution which subsequently led to a change in their logistic planning. Similar studies must be repeated in different institutions in various countries in order to find out the common obstacles as well as possible particular local solutions that speeds up the process.", "responses": [] }, { "id": "6092", "date": "10 Sep 2014", "name": "M. Hammad Ather", "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\nOperating room space and time is one of the major reasons for long waiting times in elective operative procedures. Improving operating room efficiency is therefore imperative. The authors of the current report attempted to identify lacunae in the operating room throughput and identify reasons for cancellations of elective surgeries.The authors noted that a significant proportion of time spent is non operative. This is particularly important for urology, where a typical operating list is comprised of many short endoscopic procedures. The authors recommend increasing nursing and anesthetic staff to cater to this particular need. Induction rooms are valuable as they limit the loss of non operative utilization of the OR room. It is not apparent from current rooms whether the induction room was fully utilized in the course of this work.Last minute cancellation of an operative procedure is not only an emotional and financial stress on the family/insurance but also on the hospital. Pre-operative anesthetic and nursing review is valuable in significantly reducing cancellations on the day of surgery.", "responses": [] } ]	1	https://f1000research.com/articles/3-197
https://f1000research.com/articles/3-85/v1	03 Apr 14	{ "type": "Research Article", "title": "Early elevated serum gamma glutamyl transpeptidase after liver transplantation is associated with better survival", "authors": [ "Edris M Alkozai", "Ton Lisman", "Robert J Porte", "Maarten W Nijsten", "Edris M Alkozai", "Ton Lisman", "Robert J Porte" ], "abstract": "Background: Gamma glutamyl transpeptidase (GGT) is a membrane bound enzyme that plays a key role in the synthesis of the antioxidant glutathione. Epidemiological studies have linked high GGT with an increased risk of morbidity and cardiovascular mortality. In contrast, GGT is usually elevated in liver transplant recipients that experience good outcomes.Aims: To study if and how GGT is correlated with mortality following liver transplantation.Methods: We analyzed the prognostic relevance of serum GGT levels during the early and late postoperative period after liver transplantation in 522 consecutive adults. We also studied alanine aminotransferase, aspartate aminotransferase, and total bilirubin levels.Results: Early after transplantation, the peak median (interquartile range) GGT levels were significantly higher in patients who survived more than 90 days compared to non-survivors: 293 (178-464) vs. 172 (84-239) U/l, p<0.0001. In contrast, late after transplantation, GGT levels were significantly lower in patients who survived more than 5 years than those who did not (p<0.01). The pattern of GGT levels also differed from those of alanine aminotransferase, aspartate aminotransferase, and total bilirubin early after transplantation, while these patterns were congruent late after transplantation. Kaplan-Meier survival analysis showed that early after transplantation the higher the GGT levels, the better the 90-day survival (p<0.001). In contrast, late after transplantation, higher GGT levels were associated with a lower 5-year survival (p<0.001).Conclusions: These paradoxical findings may be explained by the time-dependent role of GGT in glutathione metabolism. Immediate postoperative elevation of GGT may indicate a physiological systemic response while chronic elevation reflects a pathological response.", "keywords": [ "Gamma glutamyl transpeptidase", "alanine aminotransferase", "aspartate aminotransferase", "bilirubin", "liver transplantation", "oxidative stress", "glutathione", "survival" ], "content": "Introduction\n\nGamma glutamyl transpeptidase (GGT) is a membrane-bound enzyme that is essential for the synthesis of glutathione (GSH), a key antioxidant1. In clinical practice elevated serum GGT is generally used as an indicator of liver disease, such as biliary obstruction, alcohol consumption, and exposure to certain medical drugs1. Recently, several epidemiological studies have shown that a higher serum GGT level, even within the normal range, is associated with cardiovascular risk factors such as hypertension, hypertriglyceridemia, obesity, type 2 diabetes mellitus and stroke, as well as certain types of cancer2–10. In contrast to these studies, we observed that after surgery for ruptured abdominal aortic aneurysm11 or after liver resection12, GGT is transiently increased in patients who had a good outcome. In these short-term observational studies GGT level was inversely related to other liver laboratory parameters such as aspartate aminotransferase (ALT), alanine aminotransferase (AST) as well as total bilirubin (TBI)11,12. We observe that in the early postoperative period after a liver transplantation (LT) a transient gradual increase in GGT also occurs. How early and late postoperative changes in serum GGT are related to survival is not known.\n\nHere we present a study in which we assessed the relationship between early (postoperative day seven) elevated GGT levels with early and late survival (i.e. survival within the 90 days and five years post-LT, respectively). We also evaluated the relationship between late (six months postoperatively) elevated GGT levels with late survival.\n\nIn addition, we studied the early and late post-LT kinetics of GGT, AST, ALT, and TBI in patients who survived more than 90 days and in patients who did not. Likewise, these kinetics were compared with long-term survival.\n\n\nMaterials and methods\n\nWe conducted a single center cohort study on 522 first liver transplant patients. All LTs performed between January 1990 and August 2009 were included; excluded were pediatric patients (age <17 years), second or subsequent LT, and first LT patients who underwent a re-LT within 90 days of their first LT. Since obstructive mechanisms such as non-anastomotic stricture (NAS), acute graft rejection, and cholestatic disorders might influence GGT and TBI levels, we also specifically repeated our analysis with and without such patients. This study was performed in accordance with Dutch legislation and the local ethical committee guidelines.\n\nPatient characteristics and variables related to the perioperative management and the surgical procedure were obtained from a prospectively collected database. These included age, sex, body mass index, Karnofsky score, indication for LT, preoperative MELD (Model for End-Stage Liver Disease) score (calculated from preoperative laboratory measurements), length of hospital stay, cold ischemia time, warm ischemia time, duration of operation, combined transplant (kidney or lung), acute rejection, graft type, the number of units of allogeneic and autologous red blood cell units (RBC with 1 U containing 300 ml) and fresh frozen plasma (FFP with 1 U containing 250 ml), donor type, type of venous and bile duct anastomosis, and NAS within 90 days and within one year. When necessary, the hospital files were reviewed to complete all relevant clinical parameters.\n\nWe studied the levels of serum GGT and other liver variables postoperatively in two ways. Early postoperatively, up to postoperative day (POD) 30, we studied the levels of GGT (reference values; 0–40 U/l), ALT (0–45 U/l), AST (0–40 U/l), and TBI (0–17 µmol/l) over time in patients who survived more than 90 days after LT compared to those who did not. Late postoperatively (i.e. 90 days and beyond), we evaluated the levels of these variables at three months, six months, and one year in patients who survived more than five years compared to those patients who did not.\n\nTo evaluate the clinical relevance of early and late elevated GGT levels, we generated tertiles of low, intermediate, and high GGT levels based on equal percentiles. GGT levels at POD 7 were used to study the relationship between early elevated GGT levels and both 90-day and five-year survival. GGT levels at six months following LT were used to study the relationship of late elevated GGT with five-year survival. The last observation date for the status of patient survival for the study cohort was August 23, 2012.\n\nStatistical analyses were performed using the statistical software package SPSS 20 (IBM SPSS, Chicago, IL). Categorical variables are shown as numbers and percentages. Continuous variables are presented as means with standard deviation (SD) or as medians with interquartile range (IQR) based on their distribution. Continuous variables that were not normally distributed were compared using the Mann Whitney U test. We studied early LT mortality based on GGT levels at POD 7 as a categorical variable, using tertiles (low, intermediate, and high). Similarly, we assessed the late LT mortality based on GGT levels at six months post-LT using tertiles.\n\nPatient survival was analyzed with Kaplan-Meier analysis and the differences between the groups were assessed with the log-rank test. A p<0.05 was considered statistically significant.\n\n\nResults\n\nWe performed a total of 968 consecutive LTs in our center between January 1990 and August 2009. After excluding pediatric LTs (age <17 yr; n=290), patients who were re-transplanted within the 90 days of their first LT (n=39), second or subsequent LTs (n=101), patients with a lack of follow up data (n=11), and patients that died intraoperatively (n=5) due to brain death, cardiac failure, or uncontrollable bleeding, 522 patients were included in our analyses. The median age was 48 years (37–56), 54% of the patients were males, mean (SD) BMI was 24.8 (± 5.3), median Karnofsky score was 60 (30–70) and median MELD score was 17 (13–24) for the study population. Indications for LT were post necrotic cirrhosis (49%), cholestatic liver disease (30%), metabolic disease (10%), acute liver failure (7%), and miscellaneous (5%). Patient characteristics and the surgical variables of the entire group of 522 patients are summarized in Table 1.\n\nSome variables were not available for all patients. CIT, cold ischemia time; ICU, intensive care unit; IQR, interquartile range; INR, international normalized ratio; SD, standard deviation; WIT, warm ischemia time\n\nThe overall mortality within 90 days, one year, and five years for the study cohort was 8%, 12%, and 21%, respectively. Sepsis was the major cause of mortality (37%) within 90 days followed by multi organ failure (14%), and brain death (9%). Table 2 details all causes of mortality within the first 90 days.\n\nEarly postoperative laboratory variables are shown in Figure 1. Postoperatively, GGT levels increased gradually, reaching a maximum at POD 9 and decreased thereafter. Notably, the increase in GGT levels was significantly more pronounced, i.e. deviated more from the normal range, in patients who survived more than 90 days, as compared to those who did not: 297 (178–464) vs. 172 (69–271) U/l, p<0.0001, respectively. This pattern was different from that of postoperative levels of TBI, AST, and ALT. TBI was consistently and significantly lower in patients who survived more than 90 days following LT as compared to those who did. AST and ALT levels increased rapidly until POD 1 and POD 2, respectively, followed by rapid normalization thereafter. Contrary to GGT, the peak levels of AST and ALT were significantly lower in patients who survived more than 90 days following LT as compared to those who did not: AST 659 (326–1267) vs. 1201 (451–1990) U/l, p=0.01, and ALT 527 (280–1080) vs. 1082 (529–2631) U/l, p=0.001, respectively.\n\nCurves represent patients who survived more than 90 days (gray) and those who did not (black). Median values are shown. p < 0.05; *p ≤ 0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.\n\nThirty patients developed NAS within the 90 days post-LT. Since these patients may present with abnormal high TBI and GGT levels, we repeated the analysis with exclusion of these 30 patients with no significant affect on the GGT levels (graphical representation not shown). Also exclusion of patients who were treated for developing an acute rejection (n=103) and those who underwent LT for cholestatic liver disease (n=155) did not significantly affect the outcomes (graphical representation not shown).\n\nLate postoperative laboratory variables are shown in Figure 2. We studied the changes in GGT levels over time in patients who survived more than five years as compared to those who did not survive. Compared to patients who died within five years post-LT, those who survived more than five years had significantly lower GGT at three months post-LT; 95 (42–244) vs. 212 (92–400) U/l, p=0.001, six months post-LT; 70 (31–189) vs. 281 (103–438) U/l, p<0.001, and 1 year post-LT; 57 (25–153) vs. 124 (45–431) U/l, p=0.003, respectively. Notably, late post-LT the GGT levels showed the same patterns as TBI, AST, and ALT, i.e. higher levels in non-survivors.\n\nBars represent patients who survived more than five years (gray) and those who did not (black). Median values are shown. p < 0.05; **p ≤ 0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.\n\nWe also studied the clinical relevance of early versus late elevated GGT levels with Kaplan-Meier survival analysis. Figure 3 plots the overall 90-day and 5-year overall survival for the study cohort based on GGT-tertiles. A high GGT level at POD 7 was significantly associated with better early survival following LT (Figure 3A). The overall 90-day survival was 98% for high GGT (≥ 351 U/l), compared to 94% for intermediate GGT levels (188 and 350 U/l), and 87% for the low GGT (≤ 187 U/l), at POD 7 (Figure 3A). Similarly, five-year overall survival was 86%, 83%, and 73% for high, intermediate, and low GGT at POD 7 (p=0.003; Figure 3B), respectively. Remarkably, the differences in five-year survival mainly developed during the first three months post-LT with almost no difference in survival curves thereafter (Figure 3B). In sharp contrast with early GGT, a high GGT level six months post-LT was associated with lower five-year survival (Figure 3C). The overall survival within five years was 71% for elevated GGT (> 163 U/l), compared to 91% for intermediate GGT levels (44 and 163 U/l), and 93% for the low GGT (< 43 U/l), p<0.001.\n\nPanels A and B represent survival analysis post-LT in relation to GGT tertiles at POD 7, within 90 days and five years respectively (p=0.003). Panel C demonstrates five-year survival in relation to GGT tertiles at six months post-LT (p<0.001). Curves represent low (thick black), intermediate (thin black), and high GGT (gray).\n\n\nDiscussion\n\nIn this study, we evaluated the changes in GGT over time following liver transplantation and the clinical relevance of these changes for early and late survival. We found that a transiently elevated GGT early after LT was associated with increased survival rates within the first 90 days. In contrast, late elevation of GGT was associated with decreased five-year survival following LT. Although the early GGT elevations was also associated with five-year survival, this difference mainly developed during the first 90 days post-LT.\n\nThis peculiar effect was not observed for TBI, AST, and ALT since higher levels for these parameters at POD 7 and six months were associated with increased mortality at both 90 days and five years after LT.\n\nTo our knowledge, this is the first study showing the short and long term kinetics of GGT and the clinical relevance of an early elevated serum GGT in LT recipients. Previously, we have reported improved outcome in patients with significantly increased levels of GGT in the early post-operative period following liver resection12 and surgical repair of a ruptured abdominal aortic aneurysm11. However, those studies were not designed to address changes in GGT progression over time.\n\nWhile we acknowledge that association does not necessarily indicate causation, these data support the hypothesis that high GGT in an early post-LT setting may be a marker of some protective process.\n\nAlthough the precise mechanism responsible for an elevated serum GGT early after LT is yet to be determined, experimental studies have demonstrated that cellular GGT modulates crucial redox-sensitive functions, such as antioxidant and antitoxic defenses, cellular proliferation, and apoptotic balance13. Cellular GGT is a key enzyme in the gamma-glutamyl cycle resulting in production of intracellular GSH14–16, an important antioxidant agent that protects the cells against reactive oxygen species (ROS)17–19. GSH has been shown to protect the liver against ischemia reperfusion injury in animal models16,20,21. Hepatic ischemia can cause elevation of serum GGT with peak blood levels within 20 and 30 hours after restoration of hepatic arterial blood flow18,24. Reperfusion is associated with a surge of ROS, which may overwhelm host natural antioxidant defenses21. The oxidative stress from the ROS formed after reperfusion may lead to increased cellular death by damaging membrane lipids through peroxidation, disrupting normal enzymatic activities, and diminishing mitochondrial oxidative metabolism22. Cardin and colleagues23 studied oxidative stress in patients with chronic hepatitis C virus infection. Surprisingly, the authors observed an association between GGT and 8-hydroxydeoxyguanosine (8-OHdG), a marker of oxidative DNA damage. Patients who had a high level of 8-OHdG had significantly higher GGT levels but normal ALT levels23.\n\nThus, a transient increase in GGT level post-LT may reflect the host compensatory mechanism against oxidative stress and toxic metabolites generated by hypoxia, reperfusion, and surgical stress21. Therefore, an increased GGT early after LT may reflect the ability of the host to initiate an appropriate systemic response.\n\nAnother explanation for the elevated serum GGT in the early post-LT period that has been suggested relates to liver regeneration. Eisenbach and colleagues25, showed that an early increase in serum GGT after LT was associated with a better outcome and the authors reasoned that this rise could be due to liver regeneration. Although the liver might regenerate to some extent after LT, there is no conclusive evidence to support this hypothesis. As we mentioned earlier, we observed a transient increase in serum GGT levels in patients who survived a surgical repair for a ruptured abdominal aortic aneurysm11. In the latter group, it is less likely that significant liver regeneration occurs.\n\nContrary to early elevated GGT, but in line with published literature2–10, we observed that a late (i.e. six months post-LT) elevated GGT was significantly associated with decreased survival within the 5-years following LT. Although the finding that normal GGT levels in the late post-operative period is predictive of good outcomes is obvious and intuitive, the contrasting influences of GGT levels between early and late post-LT periods on survival may be compatible with the physiologic function of intracellular GGT. Notably, at three months, six months, and one year post-LT, the relative increase in serum GGT was two to four times higher in patients who did not survive for more than five years compared to those who did survive. This proportion was much higher than that of AST, ALT and TBI, which might imply that an elevated serum GGT is not only a marker of harm to the liver but it could be seen as a systemic response to harmful environmental factors. Indeed, in two studies, Lee and colleagues26,27 postulated that serum GGT in the general population might be a marker of increased exposure to environmental stress, internal xenobiotics, or other unknown factors that cause oxidative stress in the long run.\n\nTo avoid possible bias we performed our analysis after excluding obstructive mechanisms such as NAS, cholestatic disorders, and acute graft rejection early postoperatively. Exclusion of these cases did not change our results significantly, suggesting that the elevation in serum GGT early post-LT is independent of obstructive disorders.\n\nA practical clinical consequence of our findings may be that care providers in hospitals should realize that an abnormally high GGT early post-LT is not a cause for alarm or specific diagnostic procedures. In fact, a GGT activity four to five times above the normal range during the second post-operative week might even be considered beneficial.\n\nWe acknowledge some considerable limitations in this study. First, due to the retrospective design of the study we can only identify association rather than causation. This could only be established by specific (intervention) studies that measure the interaction between serum GGT and ROS markers post-LT. However, there is a strong correlation between GGT and oxidative DNA damage in cirrhotic patients23. Next, we cannot entirely exclude that liver regeneration plays a role in the early post-LT rise of GGT levels since GSH and to some extent GGT are mainly produced by the liver1. Hence a high serum GGT in patients who survived more than 90 days can also be a reflection of a well-functioning graft in LT patients. It will be important to understand the relationship of serum GGT and cellular GGT in the period immediately after surgery. Besides, cumulative evidence suggests that there is a relationship between the induction and release of ROS and ischemia reperfusion injury after other types of abdominal surgery18,21,24.\n\nIn conclusion, an elevated GGT level early after LT was associated with a better short-term outcome. However, chronically elevated GGT was associated with poor long-term outcome in the outpatient setting after LT. This peculiar switch in the prognostic meaning of GGT may result from the superposition of several mechanisms. Apparently, higher expression of GGT protects in the acute phase but reflects chronic damage over the long term.\n\n\nData availability\n\nfigshare: Serum gamma glutamyl transpeptidase levels post-liver transplantation and survival data, http://dx.doi.org/10.6084/m9.figshare.90034328", "appendix": "Author contributions\n\n\n\nThe study was conceived by EMA, TJL, RJP, and MWN. Data acquisition and analysis were performed by EMA and MWN. The paper was written by EMA, TJL, RJP and MWN.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was sponsored in part by a Mozaiek grant of the Dutch Organisation of Scientific Research (017.007.115) to Edris Alkozai.\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 J. T. Bottema for providing us with additional data.\n\n\nReferences\n\nWhitfield JB: Gamma glutamyl transferase. Crit Rev Clin Lab Sci. 2001; 38(4): 263–355. PubMed Abstract \| Publisher Full Text\n\nFraser A, Harris R, Sattar N, et al.: Alanine aminotransferase, gamma-glutamyltransferase, and incident diabetes: the British Women's Heart and Health Study and meta-analysis. Diabetes Care. 2009; 32(4): 741–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKazemi-Shirazi L, Endler G, Winkler S, et al.: Gamma glutamyltransferase and long-term survival: is it just the liver? Clin Chem. 2007; 53(5): 940–6. PubMed Abstract \| Publisher Full Text\n\nKristenson H, Ohrn J, Hood B: Convictions for drunkenness or drunken driving, sick absenteeism, and morbidity in middle-aged males with different levels of serum gamma-glutamyltransferase. Prev Med. 1982; 11(4): 403–16. PubMed Abstract \| Publisher Full Text\n\nStrasak AM, Kelleher CC, Klenk J, et al.: Longitudinal change in serum gamma-glutamyltransferase and cardiovascular disease mortality: a prospective population-based study in 76,113 Austrian adults. Arterioscler Thromb Vasc Biol. 2008; 28(10): 1857–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeterson B, Trell E, Kristensson H, et al.: Comparison of gamma-glutamyltransferase and other health screening tests in average middle-aged males, heavy drinkers and alcohol non-users. Scand J Clin Lab Invest. 1983; 43(2): 141–9. PubMed Abstract \| Publisher Full Text\n\nLee DS, Evans JC, Robins SJ, et al.: Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol. 2007; 27(1): 127–33. PubMed Abstract \| Publisher Full Text\n\nRuttmann E, Brant LJ, Concin H, et al.: Gamma-glutamyltransferase as a risk factor for cardiovascular disease mortality: an epidemiological investigation in a cohort of 163,944 Austrian adults. Circulation. 2005; 112(14): 2130–7. PubMed Abstract \| Publisher Full Text\n\nBreitling LP, Claessen H, Drath C, et al.: Gamma-glutamyltransferase, general and cause-specific mortality in 19,000 construction workers followed over 20 years. J Hepatol. 2011; 55(3): 594–601. PubMed Abstract \| Publisher Full Text\n\nKengne AP, Czernichow S, Stamatakis E, et al.: Gamma-glutamyltransferase and risk of cardiovascular disease mortality in people with and without diabetes: pooling of three British Health Surveys. J Hepatol. 2012; 57(5): 1083–9. PubMed Abstract \| Publisher Full Text\n\nHaveman JW, Zeebregts CJ, Verhoeven EL, et al.: Changes in laboratory values and their relationship with time after rupture of an abdominal aortic aneurysm. Surg Today. 2008; 38(12): 1091–101. PubMed Abstract \| Publisher Full Text\n\nAlkozai EM, Lisman T, Porte RJ: Bleeding in liver surgery: prevention and treatment. Clin Liver Dis. 2009; 13(1): 145–54. PubMed Abstract \| Publisher Full Text\n\nKobayashi H, Nonami T, Kurokawa T, et al.: Changes in the glutathione redox system during ischemia and reperfusion in rat liver. Scand J Gastroenterol. 1992; 27(8): 711–6. PubMed Abstract \| Publisher Full Text\n\nMoriya S, Nagata S, Yokoyama H, et al.: Expression of gamma-glutamyl transpeptidase mRNA after depletion of glutathione in rat liver. Alcohol Alcohol Suppl. 1994; 29(1): 107–11. PubMed Abstract\n\nRajpert-De Meyts E, Shi M, Chang M, et al.: Transfection with gamma-glutamyl transpeptidase enhances recovery from glutathione depletion using extracellular glutathione. Toxicol Appl Pharmacol. 1992; 114(1): 56–62. PubMed Abstract \| Publisher Full Text\n\nStein HJ, Oosthuizen MM, Hinder RA, et al.: Oxygen free radicals and glutathione in hepatic ischemia/reperfusion injury. J Surg Res. 1991; 50(4): 398–402. PubMed Abstract \| Publisher Full Text\n\nStenius U, Rubin K, Gullberg D, et al.: gamma-Glutamyltranspeptidase-positive rat hepatocytes are protected from GSH depletion, oxidative stress and reversible alterations of collagen receptors. Carcinogenesis. 1990; 11(1): 69–73. PubMed Abstract \| Publisher Full Text\n\nStastny F, Lisy V, Tomasova H, et al.: Effects of short-term and prolonged aerogenic hypoxia on gamma-glutamyl transpeptidase activity in the brain, liver, and biological fluids of young rats. Neurochem Res. 1985; 10(6): 819–28. PubMed Abstract \| Publisher Full Text\n\nGupta S, Rajvanshi P, Malhi H, et al.: Cell transplantation causes loss of gap junctions and activates GGT expression permanently in host liver. Am J Physiol Gastrointest Liver Physiol. 2000; 279(4): G815–26. PubMed Abstract\n\nSchauer RJ, Gerbes AL, Vonier D, et al.: Glutathione protects the rat liver against reperfusion injury after prolonged warm ischemia. Ann Surg. 2004; 239(2): 220–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang W, Wang M, Xie HY, et al.: Role of reactive oxygen species in mediating hepatic ischemia-reperfusion injury and its therapeutic applications in liver transplantation. Transplant Proc. 2007; 39(5): 1332–7. PubMed Abstract \| Publisher Full Text\n\nNonn L, Berggren M, Powis G: Increased expression of mitochondrial peroxiredoxin-3 (thioredoxin peroxidase-2) protects cancer cells against hypoxia and drug-induced hydrogen peroxide-dependent apoptosis. Mol Cancer Res. 2003; 1(9): 682–9. PubMed Abstract\n\nCardin R, Saccoccio G, Masutti F, et al.: DNA oxidative damage in leukocytes correlates with the severity of HCV-related liver disease: validation in an open population study. J Hepatol. 2001; 34(4): 587–92. PubMed Abstract \| Publisher Full Text\n\nLiu W, Schob O, Pugmire JE, et al.: Glycohydrolases as markers of hepatic ischemia-reperfusion injury and recovery. Hepatology. 1996; 24(1): 157–62. PubMed Abstract \| Publisher Full Text\n\nEisenbach C, Encke J, Merle U, et al.: An early increase in gamma glutamyltranspeptidase and low aspartate aminotransferase peak values are associated with superior outcomes after orthotopic liver transplantation. Transplant Proc. 2009; 41(5): 1727–30. PubMed Abstract \| Publisher Full Text\n\nLee DH, Blomhoff R, Jacobs DR Jr: Is serum gamma glutamyltransferase a marker of oxidative stress? Free Radic Res. 2004; 38(6): 535–9. PubMed Abstract \| Publisher Full Text\n\nLee DH, Gross MD, Steffes MW, et al.: Is serum gamma-glutamyltransferase a biomarker of xenobiotics, which are conjugated by glutathione? Arterioscler Thromb Vasc Biol. 2008; 28(4): e26–8; author reply e29. PubMed Abstract \| Publisher Full Text\n\nAlkozai EM, Lisman T, Porte RJ, et al.: Serum gamma glutamyl transpeptidase levels post-liver transplantation and survival data. Figshare. 2014. Data Source" }	[ { "id": "4362", "date": "22 Apr 2014", "name": "Michael Olausson", "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 very useful paper on Gamma Glutamyl Transpeptidase (GGT or just GT) - a test used in many liver transplant units and in hepatologist outpatient clinics. Its main use is the proposed correlation to liver damage - may it be cholestasis (often in combination with ALP elevation), rejection or infection. It is also used in alcoholic abuse monitoring. The early kinetics after liver transplantation are less known, and this paper gives valuable information regarding this. The authors found that livers with longterm function had an early raise in GT levels, not indicating signs of liver damage, while late raised levels showed the more expected liver damage as reason. Possible causes for early elevated levels are discussed, and a very relevant review article written by Whitfield (2001) is recommended.Title:The title reflects the take home message.Abstract:I think the abstract should stress the connection to oxidative stress mentioned in the review article more, since this seems to be a very plausible mechanism.Methods:The study is well conducted and patient selection well explained. The validity check of the findings by excluding NAS and cholestatic patients strengthen the results.Conclusions:The conclusions could merit from referring to the correlation to oxidative stress more. It is not proven that higher early GT actually protects the liver - it could be a surrogate marker for a liver in less oxidative stress.", "responses": [ { "c_id": "937", "date": "10 Aug 2014", "name": "Edris M Alkozai", "role": "Author Response", "response": "We would like to thank Prof. Olausson for his constructive and encouraging comments. We agree with Prof. Olausson that the oxidative stress might be a very plausible mechanism. As Dr. Olausson suggested, we have changed the abstract so that it now better reflects the role of GGT in oxidative stress.With regard to the conclusion in the discussion, Prof. Olausson points to an important issue. These data do not prove that the higher early GGT protects the liver. In our discussion we did mention that GGT is associated with better outcome and that this association does not necessarily indicate causation. Thus, our data can only support the hypothesis that high GGT in an early post-transplantation setting may be a marker of some protective process. Accordingly, we have changed that last sentence in our conclusion." } ] }, { "id": "5247", "date": "10 Jul 2014", "name": "Peter Jansen", "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 paper providing novel insights into the value of traditional serum markers. It is quite true that the understanding of elevated GGT values is very limited, despite the fact that this value is one of the most requested 'liver function tests'. Most clinicians who request this test probably have no clear understanding what is being tested. It is very interesting that elevated GGT shortly after liver transplantation is associated with a better outcome. The title, abstract, data section, statistics and figures of the paper are OK, the discussion section could be a little more to the point.In my opinion there might be two reasons for an elevated GGT early after transplantation: 1) Elevated GGT is associated with oxidative stress, and some oxidative stress may prompt the hepatocyte to activate innate protective mechanisms; 2) Elevated GGT after transplantation may be an indicator of good early secretory function of the liver. GGT in part comes from the surface of the bile ducts and is released from the anchor which attaches this enzyme to the cell surface, by bile acids in bile. These same authors have shown that immediately after transplantation the bile acid/phospholipid ratio in bile is elevated and (as in MDR3 deficient/LPAC patients) this leads to elevated GGT. Thus elevated GGT immediately after transplantation indicates an active bile flow. This protects hepatocytes from cytotoxic bile acids. Later after transplantation when the BA/PL ratio has normalized, bile has lost some of its detergent action, with lower basal serum GGT levels as a result. When GGT is elevated at this stage it has become the traditional indicator of liver damage associated with a bad outcome, as the authors indicate.If the authors agree with this reasoning they could include some of these thoughts in the discussion section, in order to provide a little bit more insight into why elevation under some conditions may be beneficial as this, in my opinion, is the main and novel message of this paper.", "responses": [ { "c_id": "938", "date": "10 Aug 2014", "name": "Edris M Alkozai", "role": "Author Response", "response": "We thank Prof. Jansen for reviewing our paper and for his exclusive discussion about potential mechanisms responsible for elevation of GGT early after transplantation.Prof. Jansen postulates two plausible reasons for increased GGT early after transplantation: the association of GGT with oxidative stress as we pointed out in our paper, and an indication of good early secretory function of the liver after transplantation that protects the hepatocytes from cytotoxic bile acids. Prof. Jansen also implies that the imbalance between the bile acids and phospholipids describes the underling mechanism of the damage to the hepatocytes early after transplantation, which disappears late after transplantation. We agree with Prof. Jansen and have included his valuable points into our discussion." } ] } ]	1	https://f1000research.com/articles/3-85
https://f1000research.com/articles/3-145/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "CyKEGGParser: tailoring KEGG pathways to fit into systems biology analysis workflows", "authors": [ "Lilit Nersisyan", "Ruben Samsonyan", "Arsen Arakelyan", "Lilit Nersisyan", "Ruben Samsonyan" ], "abstract": "The KEGG pathway database is a widely accepted source for biomolecular pathway maps. In this paper we present the CyKEGGParser app (http://apps.cytoscape.org/apps/cykeggparser) for Cytoscape 3 that allows manipulation with KEGG pathway maps. Along with basic functionalities for pathway retrieval, visualization and export in KGML and BioPAX formats, the app provides unique features for computer-assisted adjustment of inconsistencies in KEGG pathway KGML files and generation of tissue- and protein-protein interaction specific pathways. We demonstrate that using biological context-specific KEGG pathways created with CyKEGGParser makes systems biology analysis more sensitive and appropriate compared to original pathways.", "keywords": [ "The KEGG pathway database is a widely accepted source for biomolecular pathway maps and has long been considered as the gold standard for pathway-based analyses due to well-formatted human-readable maps supplemented with machine-readable XML files (KGML)", "quality of curation and comprehensiveness1. However", "the KEGG pathway database suffers from a number of limitations that reduce the adaptability of the pathways for automated analysis. These include inconsistencies in KGML files supplied with each pathway image", "such as absence of event or entity labels (e.g.", "links to other pathways or biological process labels)", "reversed directions for some associations", "absence of some interactions", "and inconsistent representation of compound interactions2. Additionally", "some features of KEGG pathways such as protein complex nodes and node duplication", "enhance graphical representation", "but reduce their machine-readability. Another limitation concerns abstractions (generalizations) used in pathway construction: (1) paralogous genes", "not always occurring together in the same biological context", "are grouped into single nodes", "and (2) all the genes are assumed to be expressed and present in the same pathway. Finally", "the sources of information on interactions depicted in pathways differ in quality and the nature of interactions (indirect", "physical", "regulatory", "etc.). Even accounting for these bottlenecks", "the KEGG pathway database is still a highly valued resource", "and we aimed to develop a tool that would make the best use of the information collected in it." ], "content": "Introduction\n\nThe KEGG pathway database is a widely accepted source for biomolecular pathway maps and has long been considered as the gold standard for pathway-based analyses due to well-formatted human-readable maps supplemented with machine-readable XML files (KGML), quality of curation and comprehensiveness1. However, the KEGG pathway database suffers from a number of limitations that reduce the adaptability of the pathways for automated analysis. These include inconsistencies in KGML files supplied with each pathway image, such as absence of event or entity labels (e.g., links to other pathways or biological process labels), reversed directions for some associations, absence of some interactions, and inconsistent representation of compound interactions2. Additionally, some features of KEGG pathways such as protein complex nodes and node duplication, enhance graphical representation, but reduce their machine-readability. Another limitation concerns abstractions (generalizations) used in pathway construction: (1) paralogous genes, not always occurring together in the same biological context, are grouped into single nodes, and (2) all the genes are assumed to be expressed and present in the same pathway. Finally, the sources of information on interactions depicted in pathways differ in quality and the nature of interactions (indirect, physical, regulatory, etc.). Even accounting for these bottlenecks, the KEGG pathway database is still a highly valued resource, and we aimed to develop a tool that would make the best use of the information collected in it.\n\nThere is a wide variety of software that manipulate with KEGG pathways, both standalone and Cytoscape 3 apps, such as KEGGscape (http://apps.cytoscape.org/apps/keggscape) for KEGG pathways visualization and data integration, and others. However, none of the available apps addresses inconsistencies in KGML files, and nor do they deal with abstractions of KEGG pathways. Herein, we describe CyKEGGParser app for Cytoscape 3 for KEGG pathway retrieval, visualization, adjustment for inconsistencies in computer-assisted manner, context-specific pathway generation, and exporting the pathways in KGML and BioPAX formats. CyKEGGParser is best suited for KEGG signaling pathways.\n\n\nImplementation\n\nThe software is implemented in Java and is available as an app for Cytoscape 3. The general workflow of CyKEGGParser is presented in Figure 1.\n\nThe input for parsing is KGML formatted files, either stored locally or downloaded from the web via REST-based KEGG API. The KEGG API can be used for individual downloads for academic use only; bulk download and non-acedmeic usage requires a KEGG FTP subscription and license agreement (http://www.kegg.jp/kegg/legal.html). The pathway selection dialogue provides a list of all KEGG pathways and organisms, however, if pathway KGML does not exist in the database, the user will receive a warning message.\n\nEach KGML file contains entries <pathway>, <entry> and <relation>, which are parsed using Java SAXParser API for reading XML files. The information contained in these entries is kept in Java objects which are instances of Graph, KeggNode and KeggRelation classes. These classes are implemented in CyKEGGParser and are independent of Cytoscape API. All the modifications applied by inconsistency correction algorithms are performed on these objects. Implementation of semi-automatic correction in CyKEGGParser is inherited from KEGGParser and described in detail by Arakelyan and Nersisyan2.\n\nOnce the final Graph with its nodes and edges is created, it is converted into CyNetwork, CyNode and CyEdge objects using CyKEGGParser’s KeggNetworkCreator class. During the conversion, all the attributes contained in Graph, KeggNode and KeggEdge objects are set in respective Cytoscape attribute tables. More specifically, we use default CyTables for network, nodes and edges, populating them by creating a new CyColumn for each of the attributes and setting the values in CyRows during iteration over nodes and edges. After CyNetwork, CyNode and CyEdge objects are created, the algorithm iterates through each CyNode, creating a separate view for it and assigning coordinates from respective attributes for X and Y positions. Finally, CyKEGGParser creates attribute-based “kegg_vs” visual style, which is applied on the network with VisualMappingManager of Cytoscape API. However, any Cytoscape visual style may be applied depending on the user’s choice.\n\nAll the corrections performed on the network, as well as tuning and saving steps (described below) are tracked in separate log files (see the User Manual provided in the Cytoscape Help menu and at http://molbiol.sci.am/big/apps/cy_kp/jar/CyKEGGParser_User_Manual.pdf).\n\nKEGG metabolic pathways, along with <relation> entries, which characterize protein-protein interaction networks (enzyme interactions, in this case), also contain <reaction> entries, characterizing compound interactions (chemical networks, http://www.kegg.jp/kegg/xml/docs/). Since CyKEGGParser relies on protein-protein interactions (PPI), parsing of metabolic pathways is not always as accurate as it is for signaling pathways. However, if only protein-protein interactions are of concern and if the KGML file contains respective <relation> entries, CyKEGGParser will parse metabolic pathways similar to signaling ones.\n\nAlong with the ability to modify the pathways by adding and deleting nodes and edges using Cytoscape-inherent tools, the user may as well customize (or “tune”) pathways according to specific biological context: particular tissue or cell type, and experimentally confirmed physical interactions.\n\nTissue-specific tuning. Tissue-specific tuning is aimed at providing the user with the ability to modify the networks based on genes expressed in a chosen cell/tissue type. Gene expression data for tuning is derived from BioGPS (http://biogps.org/) experiments for human normal and cancer tissues, provided by GeneCards (www.genecards.org), or may be supplied by the user (refer to User Manual for details). Along with specifying the source of data, the user chooses the tissue and specifies gene expression threshold.\n\nThe algorithm firstly clones the network preserving all the attributes, except for node and edge identifiers (those should be unique). Then it iterates over all the genes contained in the cloned network nodes, and removes the genes with expression values less than the specified threshold. If a node contains at least one gene that is expressed in current tissue, it remains in the network, otherwise it is removed. Nodes other than of type “gene” are preserved in the network.\n\nPPI based drill-down. In KEGG pathways, node entries represent groups of paralogous genes that have similar functions or interaction profiles1. The main incentive of PPI based pathway drill-down is to expand each node into its component genes and connect only those pairs of genes that have been shown to have true physical interactions. Together with tissue-specific tuning, this leads to generation of a “fine-tuned” network, in which all the components occur in the same biological context.\n\nPPI data, retrieved from the String database (http://string-db.org/), have been loaded in an internal MySQL database, located at the server of Bioinformatics Group of the Institute of Molecular Biology NAS RA (http://molbiol.sci.am/big/). The user can choose the source of interactions from the list of databases (GRID, DIP, KEGG, MINT and PDB), as well as set interaction confidence score threshold.\n\nThe algorithm initially creates a new network, copying all the nodes and node attributes from the former one. Afterwards, it drills down the new network through expanding each node of “gene” type into separate nodes for each member gene. Furthermore, the algorithm iterates over all the pairs of interacting nodes, and connects those members for which there is physical interaction in the corresponding PPI database. Attributes of newly assigned edges are copied from the former network table. After the drill down, duplicated nodes are combined into single ones, and isolated nodes are removed from the network.\n\nCyKEGGParser provides the functionality of saving the processed pathways back in valid KGML format, so that the modified pathways may be used outside of Cytoscape. All the modifications done to the network are saved in the attributes specific to KGML format. In addition, CyKEGGParser uses KEGGTranslator3 binary file, embedded in the app package, for KGML conversion to BioPAX2 and BioPAX3 formats (see User Manual for details).\n\n\nResults and discussion\n\nWe have taken KEGG B Cell Receptor Signaling Pathway as an example to demonstrate CyKEGGParser functionality and its applicability in pathway-based biology research. B cells are important players in humoral immunity, and their main function is dependent on the B Cell Receptor Signaling Pathway, which is initiated by antigen binding to B cell receptor. We have tuned the B Cell Receptor Signaling Pathway based on BioGPS tissue-specific gene expression data in CD19 B and CD4 T cells (see Implementation:, Pathway tuning section), and compared pathway topologies in each case.\n\nParsing and corrections. Figure 2 shows the pathway parsed with CyKEGGParser with automatic correction options applied. These include three cases of protein-compound-protein (PCP) interaction processing, reversing binding interaction directions of seven edges and processing of two group nodes.\n\nRed arrows indicate edges created by PCP corrections, blue double arrows indicate reversed interactions, and violet dashed rectangles indicate processed group nodes.\n\nTissue-specific tuning. We performed B Cell Receptor Signaling Pathway tuning in CD19 B cells and CD4 T cells. Gene expression threshold was set to 25 percentile of gene expression values of the BioGPS data. After tuning, from the 57 nodes available in the original pathway, 54 nodes remained in B cells and 52 nodes remained in T cells. Two nodes, namely, LYN, and CD19 are missing in the B Cell Receptor Signaling Pathway tuned in T cells (Figure 3). Due to their topological importance in signal propagation from the receptors to the target nodes, absence of these two nodes leads to almost complete deactivation of the entire pathway in T cells.\n\nTuning was performed with 25 percentile threshold of gene expression values for each tissue. Highlighted in yellow are the nodes (“LYN” and “CD19”) not present in the pathway tuned in CD4 T cells.\n\nProtein-protein interaction based tuning. The CD19 B cell tissue-specific version of the pathway was further tuned based on PPI. All the database sources (GRID, MINT, KEGG, DIP, PDB) were chosen and 0.8 confidence score threshold was set. Comparison of the PPI-tuned and the original networks showed that the node “VAV3…”, which contains three genes, VAV1, VAV2 and VAV3, was duplicated in the original pathway, but remained only in one place in the tuned network (Figure 4). Moreover, of the three VAV member genes only VAV1 interacts with CD19 and BLNK, transducing the signal to rac1 and rac2 nodes. This observation is in accordance with a previously published study indicating VAV1 as the only player in B Cell Receptor Signaling Pathway4.\n\nTissue-based tuning was performed with 25 percentile gene expression threshold. The confidence score for PPI-based tuning was set to 0.8 and all database sources were included.\n\nTo further demonstrate necessity of tissue-specific tuning for assessment of pathway activity changes, we compared pathway flows in original and tuned KEGG Calcium Signaling Pathways with three gene expression Datasets (norm vs B05 and B01) in CD14 monocytes, Adipocytes, and Cardiac myocytes (see Supplementary materials for details). For calculations, we have used the Pathway Scoring Application for Cytoscape5. The simulations show that pathway tuning increases the sensitivity of the pathway for signal flow analysis and thus the ability of the method to detect differentially expressed gene-related changes (Figure 5).\n\nTarget nodes are: M-“MAPK signaling pathway”, P-“Phosphatidylinositol signaling system”, A-“Apoptosis”. Tissues are: “CD14_mo”-CD14 monocytes, “Adip”-Adipocytes, “Card”-Cardiac myocytes. Data is shown in Data Set 1.\n\n\nConclusion\n\nWe have developed CyKEGGParser app for Cytoscape 3 that allows for import, correction, visualization, and tuning of KEGG pathways. Although KGML-based pathway import in Cytoscape has also been addressed by KGMLReader (http://apps.cytoscape.org/apps/kgmlreader) and KEGGscape (http://apps.cytoscape.org/apps/keggscape), semi-automatic correction and tuning-based enhancement of pathway specificity are unique and valuable features of CyKEGGParser. With this functionality we aim to maximize the effectiveness and sensitivity of gene expression-based systems biology analyses based on KEGG pathways.\n\n\nData availability\n\nF1000Research: Dataset 1. Simulation Data Sets for CyKEGGParser, 10.5256/f1000research.4410.d291076\n\n\nSoftware availability\n\nApp website: http://apps.cytoscape.org/apps/cykeggparser\n\nLatest source code: https://github.com/lilit-nersisyan/cykeggparser\n\nArchived source code as at the time of publication: 10.5281/Zenodo.103027\n\nLicense: GNU Public License 3.0: https://www.gnu.org/licenses/lgpl.html", "appendix": "Author contributions\n\n\n\nLN performed software design and development, testing and analyses, and manuscript preparation, RS implemented PPI database generation and integration, AA performed software design, algorithm development, and manuscript preparation. All the authors have read and approved the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was funded by research grant from the State Committee of Science of the Ministry of Education and Science of the Republic of Armenia, granted to Arsen Arakelyan (N 13Y-1F0022, PI: AA).\n\n\nAcknowledgements\n\nWe would like to acknowledge the GeneCards database for kindly providing normal and cancer tissue gene expression datasets.\n\n\nSupplementary materials\n\nWe have analyzed the KEGG Calcium Signaling Pathway with three gene expression datasets. As normal state (norm), we have taken BioGPS normal gene expression data for three tissues: CD14 monocytes, cardiac myocytes and adipocytes. For simulation of diseased states, we have taken the genes belonging to Calcium signaling pathway and randomly assigned two-fold change in a set of genes based on Bernoulli distribution with probabilities 0.5 (B05) and 0.1 (B01), respectively. In this way we have come up with one \"diseased state\" (B05) containing 50 and the other (B01) containing 8 differentially expressed genes (the data is provided in Data Set 2 “CalciumSignalingPathway_gene_expression_data.csv”).\n\nNext we have tuned the pathway in CD14 monocytes, cardiac myocytes and adipocytes. For each cell type, the pathway was tuned with an arbitrary threshold of 6.5 corresponding to 27–33 percentiles of gene expression values in the three tissues.\n\n\nReferences\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): D355–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArakelyan A, Nersisyan L: KEGGParser: parsing and editing KEGG pathway maps in Matlab. Bioinformatics. 2013; 29(4): 518–9. PubMed Abstract \| Publisher Full Text\n\nWrzodek C, Dräger A, Zell A: KEGGtranslator: visualizing and converting the KEGG PATHWAY database to various formats. Bioinformatics. 2011; 27(16): 2314–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPackard TA, Cambier JC: B lymphocyte antigen receptor signaling: initiation, amplification, and regulation. F1000Prime Rep. 2013; 5: 40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIsik Z, Ersahin T, Atalay V, et al.: A signal transduction score flow algorithm for cyclic cellular pathway analysis, which combines transcriptome and ChIP-seq data. Mol Biosyst. 2012; 8(12): 3224–31. PubMed Abstract \| Publisher Full Text\n\nNersisyan L, Samsonyan R, Arakelyan A: Simulation Data Sets for CyKEGGParser. F1000Research. 2014. Data Source\n\nNersisyan L, Samsonyan R, Arakelyan A: CyKEGGParser. ZENODO. 2014. Data Source" }	[ { "id": "5481", "date": "18 Jul 2014", "name": "Hans Binder", "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 CyKEGGParser: tailoring KEGG pathways to fit into systems biology analysis workflows by Lilit Nersisyan, Rouben Samsonyan and Arsen Arakelyan presents a comprehensive software tool allowing to edit KEGG pathways based on additional information taken, e.g., from tissue expression data and/or protein interaction networks. The functionalities are well described and illustrated using selected pathways in the context of B-cell signalling.In my opinion, the impact of the contribution is high especially for researchers who are interested in different kinds of biological networks and particularly who aim at interpreting experimental results in terms pathway activities.", "responses": [] }, { "id": "5509", "date": "06 Aug 2014", "name": "Augustin Luna", "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 Lilit Nersisyan, Rouben Samsonyan and Arsen Arakelyan present the Cytoscape plugin/app, CyKEGGParser. The tool provides the functionality to parse KEGG KGML files and edit interactions based on expression data and external protein-protein interaction networks. This tool would be of interest to any researcher wishing to overlay experimental results onto curated pathways. Issues: 1: Currently, the diagrams rendered in CyKEGGParser lack some elements of the original diagrams. Here is the link to the original \"B Cell Receptor Signaling Pathway\" http://www.genome.jp/kegg-bin/show_pathway?hsa04662The original KEGG diagram contains vertical lines indicating transmembrane proteins that are missing from the CyKEGGParser version. There is also a missing interaction between RAC1 and \"Regulation of actin cytoskeleton\". Are both of the elements missing from KGML or are there limitations in the capabilities of Cytoscape to render elements, such as the vertical membrane lines? Is there any resource (by the authors or others) that keeps track issues with the KGML files, if this is the result of missing information in the KGML? 2: In the PPI drill-down option, a confidence score threshold is provided, but it would be useful to direct readers to more information about this score and how to select appropriate values. What version of STRING do the authors use? Is this data automatically updated in their internal MySQL database? 3: Currently, there seems to be a problem with the BioPAX export. I believe the problem is in the way KEGGTranslator is called in that lacks the --format option (I found the call to KEGGTranslator in the parsing.log file; this was tested on OS X 10.8.5). If this is indeed the problem, it should be easy to fix. It would be good if the authors validated the resulting BioPAX output from their KGML edits to see if any other issues arise. http://biopax.baderlab.org/validator/check.htmlMissing ID cross-references (XRefs) is a likely error, but one that the authors might not be able to fix if this information is missing from the KGML file. 4: \"ppi\" in \"Set ppi threshold\" should be capitalized in the Protein-protein interaction Settings window of the \"Pathway tuning settings\". There are some other minor issues in the dialog. Some of the panels do not seem large enough to accommodate the presented text. On the \"Gene Expression Settings\" with no file the text appears as \"No file selecte\"CyKEGGParser seems stable and it performs the key function of helping users edit KEGG pathways, but it may require additional steps by users that want figures that mimic the aesthetics of the original KEGG diagrams (this may be unavoidable due to missing layout information in KGML and/or Cytoscape's inability to render all the KGML elements).", "responses": [ { "c_id": "935", "date": "07 Aug 2014", "name": "Lilit Nersisyan", "role": "Author Response", "response": "Thank you for reviewing our paper and your comments, the points you have mentioned are valuable for the app and comprehensiveness of the paper. Here is the detailed response and the description of changes we have made: 1: Currently, the diagrams rendered in CyKEGGParser lack some elements of the original diagrams. Here is the link to the original \"B Cell Receptor Signaling Pathway\" http://www.genome.jp/kegg-bin/show_pathway?hsa04662The original KEGG diagram contains vertical lines indicating transmembrane proteins that are missing from the CyKEGGParser version. There is also a missing interaction between RAC1 and \"Regulation of actin cytoskeleton\". Are both of the elements missing from KGML or are there limitations in the capabilities of Cytoscape to render elements, such as the vertical membrane lines? Is there any resource (by the authors or others) that keeps track issues with the KGML files, if this is the result of missing information in the KGML? KEGG pathway map images, besides containing information about pathway nodes and their interactions, are also decorated with various graphical notations for visualization of spatial distribution of pathway components (i.e. cells, cell and nuclear membranes, organelles, and structural proteins). Since a KGML file solely includes nodes, relations between genes (gene products), compounds and maps (in some metabolic pathways only) and reactions, it is not possible to reconstruct any graphical decoration from a KGML file.The same applies to present for protein – pathway and compound – pathway interactions, such is the case of RAC1 and “Regulation of actin cytoskeleton”. In some cases, pathway map nodes are presented to indicate that some nodes and edges are a part of another pathway. In such cases, pathway nodes should not be linked to any node via interactions. In other cases, however, the link from a node to pathway node indicates functional relationship, which is not included in KGML files. In these cases, users can add these interactions manually and save them back in KGML format using CyKEGGParser’s functionality. Unfortunately, there is no resource that keeps track of the elements that are missing in KGML files, nor is it possible for the app to “guess” what is missing while parsing the actual KGML file.2: In the PPI drill-down option, a confidence score threshold is provided, but it would be useful to direct readers to more information about this score and how to select appropriate values. What version of STRING do the authors use? Is this data automatically updated in their internal MySQL database? To clarify for the meaning of the confidence score, we have added the following paragraph in the User Manual and directed the user to String database’s help page for more details:“In String, the confidence score is derived by combining evidence about protein-protein interactions from various sources, adjusted for probability of randomly observing the interaction. More information about confidence score meanings and interaction sources can be found at the Help page of String database: http://string-db.org/newstring_cgi/show_info_page.pl?UserId=z7Cu7ePjQXnI&sessionId=rD_f8EsHGmAQ.” In the manuscript body, we have added the following paragraph, where we refer to the String database paper:“The user can choose the source of interactions from the list of databases (GRID, DIP, KEGG, MINT and PDB), as well as set interaction confidence score threshold, which is computed based on various evidence channels, adjusted for probability of randomly observing an interaction [3].”The local My-SQL database will be manually updated when new versions of String database are published. Currently, the String version 9.05 is used, and we are populating the database with interactions from version 9.1 at the moment. We have added a label on the tuning dialogue of the app, where the last updated version is seen, and added this information in the User Manual:“The interactions are manually updated in the local My-SQL database and the version of String used is mentioned on the Tuning dialogue.”3: Currently, there seems to be a problem with the BioPAX export. I believe the problem is in the way KEGGTranslator is called in that lacks the --format option (I found the call to KEGGTranslator in the parsing.log file; this was tested on OS X 10.8.5). If this is indeed the problem, it should be easy to fix. It would be good if the authors validated the resulting BioPAX output from their KGML edits to see if any other issues arise. http://biopax.baderlab.org/validator/check.htmlMissing ID cross-references (XRefs) is a likely error, but one that the authors might not be able to fix if this information is missing from the KGML file. CyKEGGParser calls KEGGTranslator for BioPAX2 and BioPAX3 conversion, with --format BioPAX_level2 and --format BioPAX_level3 options applied, respectively. Could you, please, send us the log file, to see why the –– format option is missing?KEGGTranslator performs a number of steps in order to retrieve the data missing in KGML files and come up with a valid BioPAX model, including completion of reactions, fixing invalid content of KGML entities and fetching cross-references, as described in http://www.biomedcentral.com/content/pdf/1752-0509-7-15.pdf. As In some cases, however, KEGGTranslator’s output is not successfully validated with BioPAX validator, mainly the BioPAX level 2 format. Since CyKEGGParser relies solely on command line calls to KEGGTranslator, there is nothing we can about these cases. However, we have tested whether the successfully validating outputs are also validated after pathway edits performed with CyKEGGParser or by the user in Cytoscape environment. Before calling KEGGTranslator, CyKEGGParser checks for absence of fields required by BioPAX2 and BioPAX3 formats, and adds default values as needed. This ensures that the edits do not induce problems for BioPAX conversion and further validation. The process of saving in KGML format and translating this format into BioPAX is described in the User Manual as follows:“KGML format saving assures that all the attributes required for BioPAX translation are available. For nodes, these are “entry: id” and “entry: type” attributes: these are assigned the default values (the next maximum id in the network and “gene” respectively). Node color and coordinates are not required for format conversion, however, if those are missing in the attributes, they will be assigned the values they have in Cytoscape. For interactions, the required attribute is the “type” attribute. If this attribute is missing in the network, it is assigned a value based on source and target node types, as follows: if both nodes are of type “gene” the interaction is assigned the type “PPrel”, if either of the nodes is of type “compound”, the type “PCrel” is assigned, otherwise the algorithm looks at the required format: in case of KGML and BioPAX_level2, it will assign “maplink” type to those interactions, where one of the nodes is of type “Map”, and will give the default value “PPrel” otherwise; in case of BioPAX_level3 the interaction will be removed from the network, since the latter format does not allow for interactions between non-physical entities.”4: \"ppi\" in \"Set ppi threshold\" should be capitalized in the Protein-protein interaction Settings window of the \"Pathway tuning settings\". There are some other minor issues in the dialog. Some of the panels do not seem large enough to accommodate the presented text. On the \"Gene Expression Settings\" with no file the text appears as \"No file selecte\". We have changed “ppi” to “PPI” in the “Set ppi threshold” of the dialogue, and added more space to file selection label in the “Gene Expression Settings” tab." } ] } ]	1	https://f1000research.com/articles/3-145
https://f1000research.com/articles/3-193/v1	12 Aug 14	{ "type": "Research Article", "title": "CaMKII binding to GluN2B is important for massed spatial learning in the Morris water maze", "authors": [ "Ivar S. Stein", "Michaela S. Donaldson", "Johannes W. Hell", "Ivar S. Stein", "Michaela S. Donaldson" ], "abstract": "Learning and memory as well as long-term potentiation (LTP) depend on Ca2+ influx through the NMDA-type glutamate receptor (NMDAR) and the resulting activation of the Ca2+ and calmodulin-dependent protein kinase (CaMKII). Ca2+ influx via the NMDAR triggers CaMKII binding to the NMDAR for enhanced CaMKII accumulation at post-synaptic sites that experience heightened activity as occurring during LTP. Previously, we generated knock-in (KI) mice in which we replaced two residues in the NMDAR GluN2B subunit to impair CaMKII binding to GluN2B. Various forms of LTP at the Schaffer collateral synapses in CA1 are reduced by 50%. Nevertheless, working memory in the win-shift 8 arm maze and learning of the Morris water maze (MWM) task was normal in the KI mice although recall of the task was impaired in these mice during the period of early memory consolidation. We now show that massed training in the MWM task within a single day resulted in impaired learning. However, learning and recall of the Barnes maze task and contextual fear conditioning over one or multiple days were surprisingly unaffected. The differences observed in the MWM compared to the Barnes maze and contextual fear conditioning suggest a differential involvement of CaMKII and the specific interaction with GluN2B, probably depending on varying degrees of stress, cognitive demand or even potentially different plasticity mechanisms associated with the diverse tasks.", "keywords": [ "The acquisition and storage of new tasks as well as the modification of already existing memories depend on the selective strengthening and weakening of synaptic interactions embedded in extensive networks of neurons (Kessels & Malinow", "2009", "Martin et al.", "2000", "Morris", "2013", "Neves et al.", "2008). Synaptic plasticity and especially long-term potentiation (LTP)", "a stable increase in synaptic strength", "are emerging as the cellular equivalent of learning and memory (Gruart et al.", "2006", "Whitlock et al.", "2006). Two key players involved in LTP are the Ca2+/calmodulin-dependent protein kinase II (CaMKII) and the NMDA-type glutamate receptor (NMDAR). Ca2+ influx through the NMDAR leads to CaMKII activation and recruitment to the postsynaptic density (PSD)", "which can be persistent and synapse specific (Otmakhov et al.", "2004", "Zhang et al.", "2008) (but see also (Rose et al.", "2009)). CaMKII binding to aa 1290–1309 on the NMDAR GluN2B subunit is required for this activity-dependent translocation (Halt et al.", "2012", "Leonard et al.", "1999", "Strack & Colbran", "1998", "Strack et al.", "2000) and is crucial for LTP (Barria & Malinow", "2005", "Halt et al.", "2012", "Zhou et al.", "2007). Stimulation of CaMKII results in its auto-phosphorylation on T286 causing a persistent Ca2+-independent activation of CaMKII (Lisman et al.", "2002). This auto-phosphorylation is increased after LTP and spatial learning in the Morris water maze (MWM) (Lengyel et al.", "2004", "Ouyang et al.", "1997", "Ouyang et al.", "1999", "Tan & Liang", "1996) and is required for effective binding of CaMKII to GluN2B (Bayer et al.", "2001", "Strack & Colbran", "1998). T286A CaMKII mutant mice exhibit no hippocampal NMDAR dependent LTP and have impaired MWM learning (Giese et al.", "1998)." ], "content": "Introduction\n\nThe acquisition and storage of new tasks as well as the modification of already existing memories depend on the selective strengthening and weakening of synaptic interactions embedded in extensive networks of neurons (Kessels & Malinow, 2009; Martin et al., 2000; Morris, 2013; Neves et al., 2008). Synaptic plasticity and especially long-term potentiation (LTP), a stable increase in synaptic strength, are emerging as the cellular equivalent of learning and memory (Gruart et al., 2006; Whitlock et al., 2006). Two key players involved in LTP are the Ca2+/calmodulin-dependent protein kinase II (CaMKII) and the NMDA-type glutamate receptor (NMDAR). Ca2+ influx through the NMDAR leads to CaMKII activation and recruitment to the postsynaptic density (PSD), which can be persistent and synapse specific (Otmakhov et al., 2004; Zhang et al., 2008) (but see also (Rose et al., 2009)). CaMKII binding to aa 1290–1309 on the NMDAR GluN2B subunit is required for this activity-dependent translocation (Halt et al., 2012; Leonard et al., 1999; Strack & Colbran, 1998; Strack et al., 2000) and is crucial for LTP (Barria & Malinow, 2005; Halt et al., 2012; Zhou et al., 2007). Stimulation of CaMKII results in its auto-phosphorylation on T286 causing a persistent Ca2+-independent activation of CaMKII (Lisman et al., 2002). This auto-phosphorylation is increased after LTP and spatial learning in the Morris water maze (MWM) (Lengyel et al., 2004; Ouyang et al., 1997; Ouyang et al., 1999; Tan & Liang, 1996) and is required for effective binding of CaMKII to GluN2B (Bayer et al., 2001; Strack & Colbran, 1998). T286A CaMKII mutant mice exhibit no hippocampal NMDAR dependent LTP and have impaired MWM learning (Giese et al., 1998).\n\nWe recently reported the specific disruption of CaMKII binding to GluN2B by point mutations in the GluN2B gene (Halt et al., 2012). The GluN2BKI mice contain two point mutations in the GluN2B C terminus (L1298A and R1300Q). Each of these mutations abolishes CaMKII binding to this site nearly completely in vitro (Strack et al., 2000). In vivo the two point mutations entirely abrogate the activity-dependent increase in the interaction with the NMDAR, show reduced hippocampal LTP by 50% and result in a MWM memory deficit, while acquisition of the MWM task and working memory as evaluated in the 8 arm win shift test remain normal (Halt et al., 2012).\n\nNow we report a specific moderate spatial learning deficit in the GluN2B knock in (KI) mice during massed 1-day training in the MWM. At the same time learning and memory recall in the Barnes maze and contextual fear conditioning are unaffected.\n\n\nMaterial and methods\n\nAll animal procedures were approved (Protocol #: 15512; PHS/NIH Assurance A3433-01) by the UC Davis Institutional Animal Care and Use Committee (IACUC) and followed NIH guidelines. All experiments were conducted with litter matched wild type (WT) and GluN2B KI mice of mixed sex (evenly distributed gender ratios) and between 2 and 4 months of age; mice within a cohort were not more than 3 weeks apart. For a more detailed characterization and description of the GluN2B KI mice (C57BL/6J background) including genotyping procedures refer to (Halt et al., 2012). The mice were on a 12 h light/dark cycle, housed individually for the behavioral experiments and acclimated for at least one week before all procedures. After the acclimation period and before the training and testing the mice were extensively handled (7–10 times for 1–2 min on different days). All trainings and testing were performed during the 12 h light cycle.\n\nThe MWM was a circular, enamel coated steel tank, 94 cm in diameter filled with water at 22–24°C. For the visible trial, a black labeled square platform (6 × 6 cm) emerging 2 cm out of the opaque (through addition of non-toxic paint) water surface was used, while the clear Plexiglas square platform (6 × 6 cm) for the training trials was submerged 2 cm below the opaque water surface. On the first day, the mice were trained in one visible trial followed by 12 consecutive training trials, while the platform was kept in a fixed position. The mice were always placed facing the wall of the pool and started randomly from three different starting positions that are equally distributed around the perimeter and are not located within the target quadrant. The mice were allowed to swim freely for 90 s to find the platform. If they failed to locate the platform within time, they were gently guided to it. After the mice climbed on the platform they were allowed to remain 30 s on the platform before they were removed from the pool and placed in their home cage for an inter-trial interval of 6–10 min. The mice were returned to the pool for a 90 s probe test without the platform 1 day and 7 days after the training day. The training session and probe trials were monitored and analyzed using the SMART (Version 2.5.19) real-time video-tracking system.\n\nThe Barnes maze was executed according to (Berta et al., 2007). In short, the maze consisted of a circular platform (92 cm diameter; opaque white) with 20 equally spaced holes (5 cm diameter; 7.5 cm between holes) and was elevated 105 cm above the floor. All holes were 2 cm away from the perimeter of the platform and a dark escape box (10 cm × 12.5 cm × 12.5 cm) was located beneath one of the holes.\n\nThe mouse was placed in a cylindrical black start chamber (10 cm × 12.5 cm × 12.5 cm) in the middle of the maze. After 10 s the chamber was lifted and the mouse was exposed to bright light (500 W). For the habituation trial on day 1 the animal was gently guided to the escape box and once it entered, the entry hole was covered and the mouse kept for 2 min in the dark escape box. Between each mouse, before the next trial, the platform was cleaned with a 10% Nolvasan solution to avoid remaining olfactory cues. During the acquisition phase, after the start chamber was lifted, the mouse was allowed to explore the maze for 3 min. The trial ended when the mouse entered the escape box or after 3 min had elapsed. Immediately after the mouse entered the escape box, the entry hole was covered and the animal stayed for 1 min in the dark escape box. If the mouse did not reach the target hole/escape box within 3 min, it was gently guided to it. After 1 min in the escape box, the mouse was placed back in its home cage until the next trial. Each mouse received 4 training trials per day with an inter-trial interval of 10–15 min on 4 consecutive days.\n\nThe probe trial was conducted on day 5, i.e., 24 h after the last training day, and on day 12. During the probe test the escape box was removed and the mouse was allowed to explore the maze for a fixed time of 90 s. The number of pokes (errors) in each hole and the latency and path length to reach the virtual target hole was measured using the SMART (Version 2.5.19) real-time video-tracking system.\n\nThe elevated plus maze (EPM) was a four arm maze with each arm measuring 30 × 5 cm and the central platform measured 5 × 5 cm (opaque white). One set of arms, opposing one another, were enclosed completely by grey side walls, 15 cm high, while the other set was open with a ledge of 0.5 cm on either side of the arms. The maze was elevated 100 cm from the floor and illuminated evenly with a 500W light. Mice were placed on the central platform, facing towards a closed arm, and allowed to freely explore the maze for 5 min. The SMART (Version 2.5.19) real-time tracking software was used to record the locomotor activity and the time spent on both the closed and the open arms during the test.\n\nContextual fear conditioning was conducted using the ‘Video Tracking of Fear Conditioning System’ and the ‘Video Freeze Software’ from MED Associates Inc. The experimentally naïve mice were placed in the conditioning chamber and received the first foot shock at the end of a 3 min habituation period. The mice remained for an additional minute in the chamber after the last foot shock was delivered. For the 5 shock conditioning protocol, the mice received an electric foot shock (0.75 mA, 1 s) at the end of the 3rd, 4th, 5th, 6th, and 7th minute. During the milder 3 shock (0.5 mA, 1 s) paradigm, the shock was received at the end of the 3rd, 4th and 5th minute. For the 4 day training protocol only one shock (0.75 mA, 1 s) per day was delivered after 2.5 min. In case of the milder 3 shock paradigm and the 4 day training protocol the mice were pre-exposed to the conditioning chamber for 3 min and 10 min respectively. For recall the mice were placed back for 5 min in the same chamber after the indicated periods of time. If the mice were exposed to a different context, the chamber geometry was changed (from square to curved wall), the lights were dimmed, the rod-flooring was covered with white linoleum, and the scent was changed from 0.01% bleach to bubble gum. Using the Video Freeze Program, the freezing time was determined (in %) and was defined as the absence of any movement, except for respiratory motion. The observed freezing response to the foot shock is correlated to the degree of learning, the strength of the aversive stimulus, and the number of presentations (Curzon et al., 2009).\n\nAll data are represented as mean ± the standard error of the mean (SEM). The data were analyzed by unpaired two-tailed t-test, one-way ANOVA or two-way ANOVA followed by the Bonferroni correction as indicated in the results, using GraphPad Prism 5 software. Statistical significance was considered if p ≤ 0.05.\n\n\nResults\n\nIn order to further dissect the memory deficit observed during the initial MWM experiments (Halt et al., 2012), two independent cohorts of GluN2B KI mice and their WT littermate controls (10 mice of each genotype per cohort) underwent a massed training protocol on a single day in the MWM. Such a compressed training protocol challenges the learning capacity within a short time period. It also has the potential to reveal deficits in early- versus late-phase consolidation as training is not protracted over 6 days as is the case for the classic MWM. The 1 day protocol thus might expose deficits in early consolidation when testing different cohorts 1 and 7 days after the training day without the complication that in the 6 days training paradigm training and consolidation processes overlap. After habituation on the first day (training day), mice underwent one visible platform trial followed by 12 consecutive training trials, in which the platform was no longer visible but kept in a fixed position. The initial trial with a marked visible platform is hippocampus-independent and was conducted to rule out changes in motivation, coordination, or sensory processing in GluN2B KI mice. The latency in reaching the visible platform was not different between the two genotypes (Figure 1A; WT: 51.48 ± 7.43 s, KI: 46.46 ± 6.70 s). In the following 12 consecutive trials with a submerged platform the KI mice slowly started to show deficits in spatial learning around trial 6, roughly the onset of learning in the WT mice (Figure 1B). The latencies of the 12 training trials are paired into blocks of two for analysis and the GluN2B KI mice display significantly increased average escape latencies during the last two trials (genotype: F1,228=2.63, p=0.1061; trial: F5,228=13.30, p<0.0001; genotype × trial F5,228=3.08, p=0.0103; Bonferroni post hoc test shows that the latency is significantly increased in the KI mice during the last two trials, Mdiff=20.30 s, 95% Cl [2.58, 38.01], p<0.05; Figure 1B; WT11–12: 30.31 ± 3.77 s, KI11–12: 50.61 ± 4.49 s).\n\n(A) WT and KI mice display no difference in the average escape latency during the visible platform trial. (B) The GluN2B KI mice show a deficit during acquisition of the hidden platform task. The latencies of the 12 training trials are split up into blocks of two and the average escape latency is significantly increased for KI versus WT mice during the last two training trials. (C, D) The covered distance over the blocks of two training trials is also different between KI and WT mice (C), while the mean velocity is not changed (D). (E, F) During the probe test 1 day after training the KI mice exhibit increased primary average escape latency (E) and show significantly fewer crossings into the target quadrant D (F), the former location of the submerged platform, compared to WT mice. (G) Both genotypes cover the same distance during the probe trial and display the same mean velocity. (H, I) The KI mice show a similar tendency during the probe test 7 days after training, with extended primary average escape latency (H) and diminished crossings into target quadrant D (I). (J) The covered distance and mean velocity is again unchanged. (A: adjacent right quadrant to D; B: quadrant opposite to quadrant D; C: adjacent left quadrant to D; platform: area of the former location of the platform). The same two independent cohorts, each consisting of 10 WT and 10 GluN2B KI mice, were tested in the Barnes maze before the MWM. Data represent Mean ± SEM.\n\nThe covered distance shows like the escape latency a significant genotype x trial interaction (Figure 1C; genotype: F1,228=0.16, p=0.687; trial: F5,228=43.34, p<0.0001; genotype × trial F5,228=3.13, p=0.0095), strengthening the mild learning deficit observed during the last training trials. No difference was observed between the mean velocity of the WT and GluN2B KI mice (Figure 1D).\n\nDuring the probe test one day after training, WT mice required significantly less time to reach the original location of the platform as reflected by faster primary escape latency (Figure 1E; WT: 38.09 ± 6.15 s, KI: 55.32 ± 6.91 s; t(38)=2.15, p<0.05). Analysis of the entries into the quadrants and the area of the former platform location also showed a significant difference between the two genotypes in their preference for the target quadrant (Figure 1F). The KI mice showed compared to WT no real preference for the target quadrant D and entered it significantly less often (Figure 1F; WTD: 9.45 ± 0.98, KID: 6.70 ± 1.22; genotype: F1,190=3.89, p≤0.05; quadrant/platform: F4,190=18.99, p<0.0001; genotype × quadrant/platform F4,190=1.46, p=0.2147; Bonferroni post hoc test shows that the WT mice search preferentially in the target quadrant compared to the KI mice, Mdiff=-2.750 entries, 95% Cl [-5.347, -0.153], p<0.05). The covered distance and mean velocity was not different between the two genotypes (Figure 1G), indicating similar activity levels.\n\nDuring the probe test 7 days after training some of the mice (independent of their genotype) were continuously floating and did not search actively for the escape platform. For the analysis of the 7 day probe trial, these floaters were neglected. The performance in the probe test 7 days after training showed a similar tendency compared to the one day test. The GluN2B KI leaned towards a longer primary escape latency (Figure 1H, left; WT: 29.50 ± 8.59 s, KI: 48.50 ± 5.98 s; t(18)=1.87, p=0.078) and tended to cross less often over into the target quadrant in search for the platform than WT mice (Figure 1I; WTD: 9.67 ± 1.49, KID: 6.46 ± 0.73; genotype: F1,90=3.12, p=0.0805; quadrant/platform: F4,90=16.69, p<0.0001; genotype × quadrant/platform F4,90=1.85, p=0.1259). Both genotypes again showed no observable difference in the covered distance and mean velocity (Figure 1J).\n\nIndividual analysis (one-way-ANOVA) of the search pattern (quadrant crossings) reveals that the WT mice preferentially search in the target quadrant during the 1 day (F3,76=11.61, p<0.0001) as well as the 7 day probe trial (F3,32=3.91, p=0.0175). The GluN2B KI mice conversely only show a slight non-significant preference for the target quadrant during the 1 day probe trial and no preference at all during the 7 day probe trial.\n\nThe Barnes maze is a less aversive spatial learning paradigm, which unlike the MWM allows the mice to move freely and explore the environment under conditions that are less stressful. It is solely based on the motivation of the mice to avoid prolonged exposure to an open area under light and to search for a dark and safe hideout, the escape box. Interestingly, the GluN2B KI showed no detectable deficits not only in learning but also in early memory (Figure 2). These findings contrast what had been seen before after spaced MWM learning experiments conducted over a period of 6 days (Halt et al., 2012). Over the 16 training trials, which were executed on 4 consecutive days (4 trials per day) the latency to locate and enter the escape box decreased equally fast for both genotypes and saturated on the fourth day (Figure 2A; WT: 22.99 ± 4.08 s, KI: 20.95 ± 3.95 s). During the probe test 1 day after the last training session there was no difference in the primary escape latency (Figure 2B, left; WT: 20.11 ± 6.02 s, GluN2BKI: 18.68 ± 4.72 s), covered distance (Figure 2B, middle; WT: 355.70 ± 37.25 cm, GluN2BKI: 426.60 ± 43.64 cm) or mean velocity (Figure 2B, right; WT: 3.95 ± 0.41 cm/s, GluN2BKI: 4.74 ± 0.48 cm/s). Both WT and KI mice searched preferentially in the target area and the surrounding holes (areas +1 and -1), indicating that they clearly remembered the former location of the escape box (Figure 2B, lower part). The overall number of zone entries and time in zones for GluN2B KI mice showed a slight increase compared to WT mice, which is probably a reflection of the statistically non-significant increased activity of KI as seen in the covered distance and mean velocity (Figure 2B). The performance in the probe test 7 days after the last training session mirrored the results of the 1 day probe test. There was no significant difference in the primary escape latency (Figure 2C, left; WT: 33.37 ± 7.98 s, KI: 26.80 ± 6.61 s), covered distance (Figure 2C, middle; WT: 265.40 ± 26.47 cm, GluN2BKI: 353.00 ± 49.42 cm) and mean velocity (Figure 2C, right; WT: 2.95 ± 0.29 cm/s, GluN2BKI: 3.92 ± 0.55 cm/s). The slightly increased activity of the GluN2B KI mice is again reflected by an increased number of zone entries and by the time spent in the zones (Figure 2C, bottom).\n\n(A) The escape latency for both genotypes decreases equally over the 16 training trials and saturates on the fourth day (left panel depicts average latencies over each of the 4 training days whereas right panel shows averages for each individual trial). (B) The primary escape latency during the probe test 1 day after the last training session is not different. KI mice search like the WT mice preferentially in and around the target area for the escape box (bottom panels). The covered distance (middle) and mean velocity (right) is slightly, but not significantly increased in KI mice. (C) The probe test 7 days after the last training session shows, like the 1 day test, no difference in the avg. primary escape latency (left) and the KI mice again exhibited the tendency to increased covered distance (middle) and mean velocity (right). Both genotypes searched in the right location in and around the target area (bottom panels). The same two independent cohorts, each consisting of 10 WT and 10 GluN2B KI mice, were tested in the Barnes maze followed by the MWM. Data are presented as Mean ± SEM.\n\nBasal anxiety levels in the GluN2B KI mice were assessed using the EPM. The EPM relies on the rodents’ innate fear of heights and open spaces and their preference for dark and enclosed rooms. Basal anxiety levels in the GluN2B KI mice were not different compared to their WT litter-matched controls. During exposure to the EPM the GluN2B KI mice spent most of the time in the closed arms, like the WT controls, and rarely stayed in the open arms or center (Figure 3A; in % of total time, WT: 92.30 ± 1.16%, KI: 88.68 ± 3.27%). Both genotypes covered comparable minimal distances in the open arms (Figure 3B; WT: 20.63 ± 3.96 cm, GluN2B KI: 31.62 ± 10.73 cm), while the WT mice had the tendency to be more active in the closed arms. This tendency is displayed in an increase in covered distance (Figure 3B; WT: 881.59 ± 86.40 cm, KI: 626.92 ± 59.24 cm; genotype: F1,68=10.17, p=0.0022; location: F3,68=120.21, p<0.0001; genotype × location F3,68=3.40, p=0.0225; Bonferroni post hoc test shows that the distance is significantly reduced in the KI in the closed arms, Mdiff=-254.7 cm, 95% Cl [-459.5, -49.49], p<0.01) and a trend towards a higher mean velocity in the closed arms (Figure 3C; WT: 3.19 ± 0.31 cm/s, GluN2B KI: 2.37 ± 0.21 cm/s), also apparent in the overall mean velocity (Figure 3C; WT: 3.34 ± 0.29 cm/s, GluN2B KI: 2.66 ± 0.18 cm/s). The observed normal basal anxiety levels concur with earlier findings in the open field analysis and the same innate fear reaction to trimethyl-thiazoline (TMT) (Halt et al., 2012).\n\n(A) The % of time the mice spent in the open and closed arms as well as the center is not different between WT and KI. (B) Both genotypes cover the same minimal distances in the open arms but WT mice cover greater distances than KI mice in the closed arms. (C) The overall mean velocity is not different between both genotypes, while the WT mice show a slightly increased mean velocity in the closed arm coinciding with the longer covered distance. One cohort of 10 WT and 9 GluN2B KI mice was tested. The same cohort was subsequently used for the milder three shock fear conditioning paradigm (Figure 5A–E). Data represent Mean ± SEM.\n\nWith basal anxiety levels not being affected in the GluN2B KI mice (Halt et al., 2012) (Figure 3), we investigated the role of the CaMKII/GluN2B interaction in contextual fear conditioning. The experimentally naïve mice were placed in the conditioning chamber and after a 3 min habituation period trained with five consecutive foot shocks (0.75 mA, 1 s duration; aversive unconditioned stimulus). The GluN2B KI mice learned to the same extent and at the same speed as their WT litter-matched controls (Figure 4A, C). A total of four independent cohorts were trained. To be able to distinguish possible differences between memory recall and consolidation, two cohorts were tested after 4 days (Figure 4B) and the other two cohorts after 14 days (Figure 4D). For the probe tests (after 4 or 14 days) the mice were placed back for 5 min into the conditioning chamber and the freezing response to the context was measured. There was no difference in the average time WT and KI mice spent freezing during either the 4 (WT: 51.11 ± 3.37%, KI: 52.23 ± 3.77%) or 14 days (WT: 70.48 ± 4.75%, KI: 67.51 ± 3.33%) probe test.\n\nFear conditioning was evaluated 4 days after training of two different cohorts of naïve litter-matched WT, heterozygous, and GluN2B KI mice (12–14 WT and KI mice, and 3–4 heterozygous mice per cohort) and 14 days after training with yet another two different cohorts of naïve litter-matched WT, heterozygous, and GluN2B KI mice (10 WT mice, 10–12 KI mice, and 5–7 heterozygous mice per cohort). (A, C) After the 3 min habituation phase, 5 shocks (0.75 mA, 1 s) were delivered at the end of 3rd, 4th, 5th, 6th and 7th minute (asterisks). The increase in the fraction of time spent freezing during conditioning is not different between the GluN2B KI, heterozygous, and WT mice. (B, D) All genotypes also display a comparable fraction of time freezing during the 4 day and the 14 day probe test. Data represent percentage freezing (Mean ± SEM).\n\nCaMKII T286A mutant mice have impaired contextual short-term memory (STM) and long-term memory (LTM) formation after a single or three tone-shock pairings, while contextual STM and LTM formation after five pairings are unaffected (Irvine et al., 2011; Irvine et al., 2005). To exclude the possibility that the GluN2B KI mice were over trained with five consecutive foot shocks (0.75 mA, 1 s duration) and potential cognitive deficits masked, we also tested two milder contextual fear conditioning paradigms.\n\nDuring the first paradigm, after two days of 3 min pre exposure to the conditioning chamber, 3 shocks (0.5 mA, 1 s) were delivered on day 3 at the end of the 3rd, 4th and 5th minute. Conditioning and memory tested after 4 days (WT: 53.82 ± 3.55%, KI: 53.55 ± 6.10%) were unaffected (Figure 5A, B). When the mice were exposed to a different context in the same chamber and room 1 day later, the time spent freezing was reduced, similar to pre-conditioning levels, and did not differ between the two genotypes (Figure 5C; WT: 13.33 ± 3.20%, KI: 18.90 ± 2.55%). Long term memory was tested with the same cohort 11 days after training (Figure 5D) and like the 4 day test showed no difference (WT: 47.63 ± 4.60%, KI: 53.08 ± 6.60%). Figure 5E depicts an overview of the experiment.\n\nOne cohort of litter-matched GluN2B KI and WT mice (9–10 mice per genotype, previously tested in the EPM) was fear conditioned with three shocks on a single day and a different cohort of naïve litter-matched GluN2B KI and WT mice (10 mice per genotype) was fear conditioned with four shocks over a period of 4 days. (A) KI and WT mice acquired to the same extent fear conditioning as indicated by the time spent freezing after three consecutive mild foot shocks (0.5 mA, 1 s) at the end of the 3rd, 4th and 5th minute (asterisks). (B, D) The time spent freezing during recall of context after 4 d and 11 d is not different between genotypes. (C) Exposure to a different context 5 days after training shows pre-conditioning freezing levels for both genotypes. (E) Summary plot of the experimental data from A-D. (F) Summary plot of the experimental data of the 4 day training (1 shock, 0.75 mA, 1 s per day) experiment. WT and KI mice show comparable learning curves during the 4 d conditioning and similar contextual memory when tested 6 and 8 days later. Exposure to a different context in the same room and chamber, 7 d after the last training day, revealed again pre-conditioning freezing levels. Data represent percentage freezing (Mean ± SEM).\n\nIn addition, another cohort of GluN2B KI and WT mice were trained over multiple days. The mice were pre exposed to the context on day 1 for 10 min and trained on the following 4 days with one foot shock (each 0.75 mA, 1 s) per day (Figure 5F). Again no learning or memory deficit, which was tested 6 (WT: 75.52 ± 4.64%, GluN2B KI: 62.69 ± 6.21) and 8 days (WT: 64.75 ± 5.68%, KI: 59.06 ± 7.56%) after training, was observed. When the mice were exposed to a different context in the same chamber and room on day 12, 7 days after the last training, both genotypes exhibited reduced freezing levels (WT: 25.56 ± 5.85%, KI: 22.09 ± 6.93%) similar to those observed before conditioning.\n\n\nDiscussion\n\nCaMKII and the NMDAR are crucial not only for hippocampal LTP but also for hippocampus dependent spatial learning and memory formation. In a number of studies CaMKIIα KO and mutant mice displayed learning deficits, notably mainly in aversively motivated tasks (Elgersma et al., 2002; Giese et al., 1998; Irvine et al., 2005; Silva et al., 1992), while appetitive learning was unaffected (Carvalho et al., 2001). Moreover, the regulation of expression or activity level of CaMKIIα seems to have an emotional component and especially impact emotional and anxiety-like behavior and learning. CaMKIIα heterozygous KO mice show a decreased freezing in fear conditioning induced by an electric foot shock and are more active in the open field suggesting decreased anxiety-related behaviors (Chen et al., 1994). Consistently, transgenic mice overexpressing CaMKIIα exhibit an increase in anxiety-like behaviors in open field, light-dark transition, and elevated zero maze (Hasegawa et al., 2009). CaMKIIα T286A mutant mice, which are deficient in autonomous and thereby overall CaMKII activity, spend more time in and entered more often the open arm in the EPM and generally react with hyperactivity to novel stimuli that could be perceived as potentially threatening (Easton et al., 2011). However, the behavior in the non-threatening neutral environment of the home cage is normal (Easton et al., 2011), in accordance with an impaired learning in MWM and fear conditioning (Giese et al., 1998; Irvine et al., 2005) but a normal appetitive instrumental conditioning (Carvalho et al., 2001). Results in our study concur with these findings and argue for the specific importance of CaMKII binding to GluN2B in aversive and more demanding spatial learning task of the MWM, with the less threatening and stressful Barnes maze task being unaffected in the KI mice. Contextual fear conditioning and memory was surprisingly not affected in the GluN2B KI mice. We assume that this discrepancy is due to the higher demands of the MWM task and the differential regulation of spatial and contextual LTM formation (Mizuno & Giese, 2005).\n\nLearning during massed training sessions, i.e., when training trials are given in short order on a single day, generally results in lower levels of learning than spaced training over a period of several days. For instance, in the MWM rats showed faster and better acquisition during spaced (16 trials over 4 days) than massed training (16 trials in 1 day) as well as better memory formation (Commins et al., 2003). During contextual fear conditioning 1 h spaced training also results in better memory formation than massed training, while cued fear conditioning is unaffected (Scharf et al., 2002). Likely because of the more demanding task of massed training, which requires information being stored during a shorter time period than in spaced training, the GluN2B KI mice revealed a mild learning deficit compared to WT mice in the 1 day paradigm (Figure 1) when they did not in the 6 day MWM training protocol (Halt et al., 2012). Figure 1D shows that there is no difference in the swim speed between WT and KI mice, which together with the increased covered distance of the KI mice during the last two trials (Figure 1C) argues against the differences in agility and fatigue in the GluN2B KI. The increased covered distance shows that the KI mice have to swim longer and cover a greater distance in order to find the platform during the later trials, further supporting the extended primary escape latency (Figure 1B) and therefore the KI learning deficit. The learning deficit in the 1 day paradigm suggests that the learning capacity over a more limited time period is reduced in GluN2B KI mice. In addition, the elevated stress level during the massed training in the 1 day protocol could affect learning in the GluN2B KI mice more than in WT contributing to the differential outcome.\n\nThe GluN2B KI phenotype is similar to alpha and delta CREB KO mice, which are impaired in massed but not in spaced MWM learning and contextual fear conditioning (Kogan et al., 1997). The improvement in memory after spaced learning is also mimicked in acute hippocampal slices where successive spaced theta burst stimulation resulted in enhanced previously saturated LTP (Kramar et al., 2012). In the study by Kramar et al. the induction of LTP in a subset of synapses during the first theta burst train primed the initially unresponsive neighbors (probably in a translation-dependent manner, since the two stimulations have to be spaced 1 h apart; see also (Scharf et al., 2002)), resulting in potentiation after stimulation with the second theta burst train. This mechanism will probably be especially important in cases of impaired LTP induction or expression like in our GluN2B KI mice where various forms of LTP were reduced by ~50% (Halt et al., 2012).\n\nThe Barnes maze is a demanding spatial task using visual cues in the room for navigation, which is in that regard similar to the MWM. One major difference between these two spatial learning paradigms is probably the amount of stress to which animals are exposed. The unaffected performance of the GluN2B KI mice in the less stressful Barnes Maze is in line with the known importance of CaMKII selectively in aversive rather than appetitive motivated spatial and contextual learning paradigms (Elgersma et al., 2002; Giese et al., 1998; Irvine et al., 2005; Silva et al., 1992). However, further potential differences include sensory processes and response requirements, which cannot be excluded as a reason for the deficit in massed MWM learning compared to no deficit in the Barnes maze (Figure 2).\n\nFear conditioning, on the other hand, creates a high level of anxiety. Compared to spatial learning paradigms like the MWM, the contextual conditioning only requires memorization of a far less complex environment, which only has to be recognized and does not require navigation and localization of a certain target. Additionally, at the beginning of the conditioning the mice have time to explore and form a spatial map without being stressed or more anxious than during the normal exposure to a novel environment, before the first shock is delivered at the end of the 3rd minute. During the milder conditioning paradigm (Figure 5) the mice were pre-exposed to the shocking chamber on the previous day, which allowed the formation of an even better spatial representation (Fanselow, 2000; Frankland et al., 2004; Hu et al., 2007). This lack of a demanding spatial learning requirement and possible formation of contextual memory during still low stress levels could explain the normal conditioning and contextual memory (Figure 4 and Figure 5) compared to the moderate learning deficit observed during massed MWM training (Figure 1). In addition, fear conditioning in general constitutes a very strong learning paradigm, which might override and mask any mild learning deficits because of its vigor even in the mildest form tested here. In support of this idea the five tone-shock pairings allow full fear conditioning in CaMKII T286A mutant mice when one and three pairings do not (Irvine et al., 2011; Irvine et al., 2005). Accordingly, the strength of conditioning can mask learning deficits, and for GluN2B KI mice that might even be the case for single shock experiments. Spatial and contextual LTM formation is differently regulated. There are not only divergent regional requirements within the hippocampus, but also the control of gene transcription varies between spatial and contextual LTM formation (Mizuno & Giese, 2005). Accordingly, MWM and fear conditioning could depend on molecular mechanisms that vary in detail even though both learning paradigms require CaMKII.\n\nContrasting CaMKIIα heterozygous KO, T286A KI mice, or transgenic mice overexpressing CaMKIIα, the GluN2B KI mice display no change in basal anxiety levels or anxiety-like behaviors. They show avoidance of the open arms in the EPM (Figure 3), behavior in the open field (Halt et al., 2012), and reaction to TMT (the anxiogenic compound in fox urine) (Halt et al., 2012), a measure of innate fear, that are comparable to WT mice. Thus it appears that CaMKIIα affects fear behavior independent of its binding to GluN2B.\n\nOur study suggests a specific requirement for the activity-dependent interaction of CaMKII with the NMDAR during the acquisition of more elaborate spatial learning tasks under modestly aversive, stressful conditions. The results argue not only for a role of CaMKII binding to GluN2B during consolidation (Halt et al., 2012), but also for learning, which becomes obvious under the here applied more demanding MWM conditions. The less stressful Barnes maze neither results in a consolidation deficit nor is the learning itself affected. In addition, the less demanding but more potent contextual fear conditioning and memory was surprisingly normal. Whether reduced recall of the MWM task 1 and 7 days after single day training reflects reduced learning, reduced memory consolidation, or both cannot be answered at this point.\n\n\nData availability\n\nF1000Research: Dataset 1. Data sets for behavioral tasks in wild type and GlunN2BKI mice, 10.5256/f1000research.4660.d34165 (Stein et al., 2014).", "appendix": "Author contributions\n\n\n\nISS and JWH designed the experiments. ISS with help from MSD carried out the experiments. ISS and JWH prepared the manuscript. All authors approved 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 was supported by NIH grants NS046450 and NS078792 to JWH.\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 the Personnel of the UC Davis Mouse Behavioral Assessment Lab, where all behavioral tests were conducted, for their assistance and help with data analysis, as well as Claire M. Koenig for training in the Barnes maze. We thank Dr. Karl-Peter Giese (Kings College, London, UK) for reading an earlier version of the manuscript.\n\n\nReferences\n\nBarria A, Malinow R: NMDA receptor subunit composition controls synaptic plasticity by regulating binding to CaMKII. Neuron. 2005; 48(2): 289–301. PubMed Abstract \| Publisher Full Text\n\nBayer KU, De Koninck P, Leonard AS, et al.: Interaction with the NMDA receptor locks CaMKII in an active conformation. Nature. 2001; 411(6839): 801–805. PubMed Abstract \| Publisher Full Text\n\nBerta S, Gert L, Harald H, et al.: Barnes maze, a useful task to assess spatial reference memory in the mice. Protoc exch. 2007. Publisher Full Text\n\nCarvalho MO, Silva JA, Balleine WB: Evidence of selective learning deficits on tests of pavlovian and instrumental conditioning in α-CaMKII T286 mutant mice. Human Sciences Press, New York, NY ETATS-UNIS. Int J Comp Psychol. 2001; 14(3): 161–174. Reference Source\n\nChen C, Rainnie DG, Greene RW, et al.: Abnormal fear response and aggressive behavior in mutant mice deficient for alpha-calcium-calmodulin kinase II. Science. 1994; 266(5183): 291–294. PubMed Abstract \| Publisher Full Text\n\nCommins S, Cunningham L, Harvey D, et al.: Massed but not spaced training impairs spatial memory. Behav Brain Res. 2003; 139(1–2): 215–223. PubMed Abstract \| Publisher Full Text\n\nCurzon P, Rustay NR, Browman KE: Cued and Contextual Fear Conditioning for Rodents. In Buccafusco, J.J. (ed), Boca Raton (FL). Meth Behav Anal Neurosci. 2009; 19–37. PubMed Abstract \| Publisher Full Text\n\nEaston AC, Lucchesi W, Schumann G, et al.: αCaMKII autophosphorylation controls exploratory activity to threatening novel stimuli. Neuropharmacology. 2011; 61(8): 1424–1431. PubMed Abstract \| Publisher Full Text\n\nElgersma Y, Fedorov NB, Ikonen S, et al.: Inhibitory autophosphorylation of CaMKII controls PSD association, plasticity, and learning. Neuron. 2002; 36(3): 493–505. PubMed Abstract \| Publisher Full Text\n\nFanselow MS: Contextual fear, gestalt memories, and the hippocampus. Behav Brain Res. 2000; 110(1–2): 73–81. PubMed Abstract \| Publisher Full Text\n\nFrankland PW, Josselyn SA, Anagnostaras SG, et al.: Consolidation of CS and US representations in associative fear conditioning. Hippocampus. 2004; 14(5): 557–569. PubMed Abstract \| Publisher Full Text\n\nGiese KP, Fedorov NB, Filipkowski RK, et al.: Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in LTP and learning. Science. 1998; 279(5352): 870–873. PubMed Abstract \| Publisher Full Text\n\nGruart A, Muñoz MD, Delgado-Garcia JM: Involvement of the CA3-CA1 synapse in the acquisition of associative learning in behaving mice. J Neurosci. 2006; 26(4): 1077–1087. PubMed Abstract \| Publisher Full Text\n\nHalt AR, Dallapiazza RF, Zhou Y, et al.: CaMKII binding to GluN2B is critical during memory consolidation. EMBO J. 2012; 31(5): 1203–1216. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHasegawa S, Furuichi T, Yoshida T, et al.: Transgenic up-regulation of alpha-CaMKII in forebrain leads to increased anxiety-like behaviors and aggression. Mol Brain. 2009; 2: 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu H, Real E, Takamiya K, et al.: Emotion enhances learning via norepinephrine regulation of AMPA-receptor trafficking. Cell. 2007; 131(1): 160–173. PubMed Abstract \| Publisher Full Text\n\nIrvine EE, Danhiez A, Radwanska K, et al.: Properties of contextual memory formed in the absence of αCaMKII autophosphorylation. Mol Brain. 2011; 4: 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIrvine EE, Vernon J, Giese KP: AlphaCaMKII autophosphorylation contributes to rapid learning but is not necessary for memory. Nat Neurosci. 2005; 8(4): 411–412. PubMed Abstract \| Publisher Full Text\n\nKessels HW, Malinow R: Synaptic AMPA receptor plasticity and behavior. Neuron. 2009; 61(3): 340–350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKogan JH, Frankland PW, Blendy JA, et al.: Spaced training induces normal long-term memory in CREB mutant mice. Curr Biol. 1997; 7(1): 1–11. PubMed Abstract \| Publisher Full Text\n\nKramár EA, Babayan AH, Gavin CF, et al.: Synaptic evidence for the efficacy of spaced learning. Proc Natl Acad Sci U S A. 2012; 109(13): 5121–5126. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLengyel I, Voss K, Cammarota M, et al.: Autonomous activity of CaMKII is only transiently increased following the induction of long-term potentiation in the rat hippocampus. Eur J Neurosci. 2004; 20(11): 3063–3072. PubMed Abstract \| Publisher Full Text\n\nLeonard AS, Lim IA, Hemsworth DE, et al.: Calcium/calmodulin-dependent protein kinase II is associated with the N-methyl-D-aspartate receptor. Proc Natl Acad Sci U S A. 1999; 96(6): 3239–3244. PubMed Abstract \| Publisher Full Text\n\nLisman J, Schulman H, Cline H: The molecular basis of CaMKII function in synaptic and behavioural memory. Nat Rev Neurosci. 2002; 3(3): 175–190. PubMed Abstract \| Publisher Full Text\n\nMartin SJ, Grimwood PD, Morris RG: Synaptic plasticity and memory: an evaluation of the hypothesis. Annu Rev Neurosci. 2000; 23: 649–711. PubMed Abstract \| Publisher Full Text\n\nMizuno K, Giese KP: Hippocampus-dependent memory formation: do memory type-specific mechanisms exist? J Pharmacol Sci. 2005; 98(3): 191–197. PubMed Abstract \| Publisher Full Text\n\nMorris RG: NMDA receptors and memory encoding. Neuropharmacology. 2013; 74: 32–40. PubMed Abstract \| Publisher Full Text\n\nNeves G, Cooke SF, Bliss TV: Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. Nat Rev Neurosci. 2008; 9(1): 65–75. PubMed Abstract \| Publisher Full Text\n\nOtmakhov N, Tao-Cheng JH, Carpenter S, et al.: Persistent accumulation of calcium/calmodulin-dependent protein kinase II in dendritic spines after induction of NMDA receptor-dependent chemical long-term potentiation. J Neurosci. 2004; 24(42): 9324–9331. PubMed Abstract \| Publisher Full Text\n\nOuyang Y, Kantor D, Harris KM, et al.: Visualization of the distribution of autophosphorylated calcium/calmodulin-dependent protein kinase II after tetanic stimulation in the CA1 area of the hippocampus. J Neurosci. 1997; 17(14): 5416–5427. PubMed Abstract\n\nOuyang Y, Rosenstein A, Kreiman G, et al.: Tetanic stimulation leads to increased accumulation of Ca(2+)/calmodulin-dependent protein kinase II via dendritic protein synthesis in hippocampal neurons. J Neurosci. 1999; 19(18): 7823–7833. PubMed Abstract\n\nRose J, Jin SX, Craig AM: Heterosynaptic molecular dynamics: locally induced propagating synaptic accumulation of CaM kinase II. Neuron. 2009; 61(3): 351–358. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScharf MT, Woo NH, Lattal KM, et al.: Protein synthesis is required for the enhancement of long-term potentiation and long-term memory by spaced training. J Neurophysiol. 2002; 87(6): 2770–2777. PubMed Abstract\n\nSilva AJ, Paylor R, Wehner JM, et al.: Impaired spatial learning in alpha-calcium-calmodulin kinase II mutant mice. Science. 1992; 257(5067): 206–211. PubMed Abstract \| Publisher Full Text\n\nStein IS, Donaldson MS, Hell JW: Data sets for behavioral tasks in wild type and GlunN2BKI mice. F1000Research. 2014. Data Source\n\nStrack S, Colbran RJ: Autophosphorylation-dependent targeting of calcium/calmodulin-dependent protein kinase II by the NR2B subunit of the N-methyl- D-aspartate receptor. J Biol Chem. 1998; 273(33): 20689–20692. PubMed Abstract \| Publisher Full Text\n\nStrack S, McNeill RB, Colbran RJ: Mechanism and regulation of calcium/calmodulin-dependent protein kinase II targeting to the NR2B subunit of the N-methyl-D-aspartate receptor. J Biol Chem. 2000; 275(31): 23798–23806. PubMed Abstract \| Publisher Full Text\n\nTan SE, Liang KC: Spatial learning alters hippocampal calcium/calmodulin-dependent protein kinase II activity in rats. Brain Res. 1996; 711(1–2): 234–240. PubMed Abstract \| Publisher Full Text\n\nWhitlock JR, Heynen AJ, Shuler MG, et al.: Learning induces long-term potentiation in the hippocampus. Science. 2006; 313(5790): 1093–1097. PubMed Abstract \| Publisher Full Text\n\nZhang YP, Holbro N, Oertner TG: Optical induction of plasticity at single synapses reveals input-specific accumulation of alphaCaMKII. Proc Natl Acad Sci U S A. 2008; 105(33): 12039–12044. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhou Y, Takahashi E, Li W, et al.: Interactions between the NR2B receptor and CaMKII modulate synaptic plasticity and spatial learning. J Neurosci. 2007; 27(50): 13843–13853. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5845", "date": "22 Aug 2014", "name": "Mark L. Dell’Acqua", "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 article that provides additional, detailed behavioral analysis of hippocampal-dependent spatial and contextual learning and memory performance of knock-in mice that are deficient in CaMKII binding to the GluN2B subunit of the NMDA-subtype ionotropic glutamate receptor. The current findings extend those of the authors in Halt et al. (2012) where spaced training over 6 days in the Morris Water Maze (MWM) revealed no deficits in spatial learning but did find significant deficits in consolidation of long-term memory probed several days after training in theses CaMKII binding-deficient GluN2B knock-in mice.Interestingly, here, using a more intense and demanding one day, massed training version of the MWM, the authors now detect modest deficits in these GluN2B knock-in mice, not only in memory probed one day after training, but also in spatial learning measured on the day of training. However, other single day training spatial/contextual learning and memory tests, that are both more aversive but less spatially challenging (contextual fear conditioning) and less aversive but more spatially challenging (Barnes maze), did not reveal any difference in performance between WT and GluN2B knock-in mice. Thus, an important conclusion of this study is that increased postsynaptic targeting of CaMKII and the increased level of LTP (~50%; see Halt et al.) that are achieved through CaMKII binding to GluN2B are only required for spatial learning when the task is both fairly demanding and at least moderately aversive. These findings should give the field a heightened appreciation that highly-compartmentalized regulation of signaling pathways by postsynaptic protein-protein scaffolding interactions, such as achieved through CaMKII binding to GluN2B during LTP-induction, may not be absolutely essential for spatial learning and memory in general but may have evolved to allow animals to specifically sustain optimal spatial learning and memory capabilities even under demanding and stressful circumstances.Overall, the manuscript is very well written and the data are rigorously analyzed and presented clearly.", "responses": [] }, { "id": "5919", "date": "26 Aug 2014", "name": "Karl-Ulrich Bayer", "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\nSummary: This study substantially extends the previous behavioral analysis of a mouse line with mutant GluN2B that is incapable of CaMKII binding. Individually, both GluN2B and CaMKII are critically involved in regulating synaptic plasticity and learning and memory. Previous studies from the same lab showed that the mutant GluN2B KI mice have impaired hippocampal CA1 LTP (reduced to ~50%) and impaired recall during the consolidation phase of a Morris water maze (MWM) task 1. Here, learning impairments were found in a modified MWM task that involves massed learning during day 1. All other learning paradigms tested here or previously were normal. Basal behavior was also extensively evaluated, and was found to be normal in all tasks tested here or previously. Evaluation: The manuscript is very well written. The experiments and the analyses are solid. The interpretation and discussion of the results are both careful and insightful. Overall, it can be very reasonably concluded that learning requires the CaMKII/GluN2B interaction only when the learning task is rather difficult and/or involves some stress (such as in a water maze), consistent also with impaired but not abolished LTP in these mice. Comments: The conclusions from this study may also inform potential future behavioral studies on these mice. For instance, while the GluN2B KI mice were normal in multiple variations of contextual fear conditioning, the contextual part of the task was made rather easy in all paradigms tested. However, even with shorter exposure to the context, fear conditioning may still be too easy of a learning task to reveal any differences in the GluN2B mice. Thus, while such future studies may be interesting, either outcome would still be consistent with the conclusions reached here.The behavioral analyses of the GluN2B KI mice were done specifically in order to evaluate the impact of the LTP impairments in these mice. When comparing the GluN2B KI mice with CaMKII T286A or CaMKII knock-out mice in other behavioral tasks in future studies, it may have to be taken into account that the GluN2B KI mice are impaired only in LTP, while the CaMKII mutant mice are additionally impaired in LTD. One point in which my opinion differs is within the introduction, specifically regarding the statement that “auto-phosphorylation on T286 is required for effective binding of CaMKII to GluN2B”. The cited references could indeed be interpreted support this statement2 3. However, newer data indicate that the increased binding observed after ATP addition was largely caused by direct nucleotide binding to the kinase, rather than by auto-phosphorylation4 5 6.Notably, this point does not affect the validity of any conclusions reached in the manuscript. However, it may significantly mislead the reader about the larger context of CaMKII in synaptic plasticity. Recently, we have shown that CaMKII T286 phosphorylation is required not only for LTP, but also for LTD7. Prior to this, other recent evidence had already suggested a more prolonged T286 phosphorylation after LTD than after LTP8 9 10 (for review see Coultrap & Bayer, 201211). If T286 phosphorylation would induce the most effective CaMKII binding to GluN2B (as stated), binding would be expected to be more extensive after LTD compared to LTP. However, this does not appear to be the case 9, and the CaMKII/GluN2B interaction contributes exclusively to the CaMKII functions in LTP but not in LTD1.", "responses": [] }, { "id": "6236", "date": "29 Sep 2014", "name": "Isabel Pérez-Otaño", "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 Stein et al. expands previous work from the same group characterizing the learning abilities of CamKII knock-in mice (Halt et al., 2012). The mice carry two point mutations that impair the binding of CamKII to the NMDA receptor subunit GluN2B, and were shown to exhibit deficits in memory recall in the Morris Water maze (MWM). Here, the authors present an extended and carefully performed behavioral analysis. Their new results show that spatial learning in the MWM is impaired when 12 training trials are compressed in a single day. In contrast, the previous study showed that spatial learning was normal when training was conducted over 6 consecutive days (2 trials per day). The different outcome is attributed to the increased cognitive demands and higher stress levels posed by massed training. The paper is well-written and clearly presented, and the Methods are described with enough detail. Comments: The deficits in massed spatial learning are only detectable in the last two trials. Do the authors have an explanation? Could the delay be related to a defect in an early phase of consolidation, in line with the hypothesis previously proposed by the authors (Halt et al., 2012)? It might be nice to see whether memory consolidation is impaired in other hippocampal-dependent tests not involving stress. For instance, the novel object recognition test that can be performed in a brief period of time and would additionally permits the assessment of memory recall at shorter time intervals after initial learning.", "responses": [] } ]	1	https://f1000research.com/articles/3-193
https://f1000research.com/articles/3-191/v1	12 Aug 14	{ "type": "Research Article", "title": "SkateBase, an elasmobranch genome project and collection of molecular resources for chondrichthyan fishes", "authors": [ "Jennifer Wyffels", "Benjamin L. King", "James Vincent", "Chuming Chen", "Cathy H. Wu", "Shawn W. Polson", "Benjamin L. King", "James Vincent", "Chuming Chen", "Cathy H. Wu", "Shawn W. Polson" ], "abstract": "Chondrichthyan fishes are a diverse class of gnathostomes that provide a valuable perspective on fundamental characteristics shared by all jawed and limbed vertebrates. Studies of phylogeny, species diversity, population structure, conservation, and physiology are accelerated by genomic, transcriptomic and protein sequence data. These data are widely available for many sarcopterygii (coelacanth, lungfish and tetrapods) and actinoptergii (ray-finned fish including teleosts) taxa, but limited for chondrichthyan fishes. In this study, we summarize available data for chondrichthyes and describe resources for one of the largest projects to characterize one of these fish, Leucoraja erinacea, the little skate. SkateBase (http://skatebase.org) serves as the skate genome project portal linking data, research tools, and teaching resources.", "keywords": [ "Chondrichthyan fishes are composed of two subclasses", "Holocephali and Elasmobranchii. Holocephalans are the more basal of the pair having first appeared more than 400 million years ago and include a single surviving order", "Chimaeriformes", "the chimaeras", "with 39 extant species1. Elasmobranchs appeared approximately 350 million years ago and include more than 1000 species of sharks", "skates", "and rays2. Chondrichthyan fishes occupy a pivotal position at the base of the vertebrate phylogenetic tree. For research that includes an evolutionary component", "representation of this diverse class affords a valuable perspective to evaluate all vertebrates." ], "content": "Introduction\n\nChondrichthyan fishes are composed of two subclasses, Holocephali and Elasmobranchii. Holocephalans are the more basal of the pair having first appeared more than 400 million years ago and include a single surviving order, Chimaeriformes, the chimaeras, with 39 extant species1. Elasmobranchs appeared approximately 350 million years ago and include more than 1000 species of sharks, skates, and rays2. Chondrichthyan fishes occupy a pivotal position at the base of the vertebrate phylogenetic tree. For research that includes an evolutionary component, representation of this diverse class affords a valuable perspective to evaluate all vertebrates.\n\nChondrichthyan fishes are circumglobal in distribution and occupy a wide range of ecological habitats. Their life history parameters are equally disparate but in general chondrichthyans are slow growing and late maturing fishes with an increased risk of extinction3–5. Fecundity is as few as 1 or 2 for viviparous species such as the sand tiger shark, Carcharias taurus6 and as high as 300 for the whale shark, Rhincodon typus7. They are of economic importance for fisheries as well as ecotourism. Management and assessment of stock is essential to ensure both ecotourism interests and food resources remain sustainable8. Management of fish populations has increasingly relied on molecular tools to investigate population structure, properly identify species, and compliance with fishing quotas9–13.\n\nElasmobranchs have been used as a model for biomedical research for more than 100 years. Elasmobranchs, like other cartilaginous fishes, exhibit many fundamental vertebrate characteristics, including a neural crest, jaws and teeth, an adaptive immune system, and a pressurized circulatory system. The skate is a powerful comparative model to study biological processes shared among jawed and limbed vertebrates such as development14–16, renal physiology17–20, immunology21–26, toxicology27, neurobiology28, and wound healing and regeneration29. They are the most ancient vertebrates to posses an adaptive immune system that generates antibodies using a V(D)J combinatorial mechanism30. Phylogenetically, cartilaginous fishes are the first vertebrates to possess a thymus, a central lymphoid organ that provides a microenvironment for the development of T cells31. The thymus shares a common organization with more derived vertebrates containing cortical and medullary regions32,33.\n\nIn addition to shared physiological characteristics, the diversity of specializations between species allows investigations of evolution within a single clade. For example, elasmobranchs use a plethora of reproductive strategies that span the full range of maternal investment from placental viviparity to strict lecitrophic oviparity. Besides sexual reproduction, captive elasmobranchs are capable of asexual parthenogenesis34–36. Of these reproductive mechanisms, the most tractable for research purposes is oviparity. Approximately 43% of chondrichthyans utilize oviparity including all Chimaeriformes, Heterodontiformes (bullhead sharks), Rajoidae (skates) and Scyliorhinidae (catsharks)37. Many species can be maintained in captivity and will breed and lay eggs throughout an annual season38. Artificial insemination has been reported for two oviparous species, the clearnose skate, Raja eglanteria39, and the cloudy catshark, Scyliorhinus torazame40. Additionally, sperm storage allows wild caught females to lay eggs for several years without requiring males or captive mating events41.\n\nLeucoraja erinacea, the little skate, was chosen for a genome sequencing project to represent this clade of fishes because of their use as a biomedical model, experimental tractability, genome size, existing sequence data, and northeast regional distribution. The sequencing project is an ongoing effort of the North East Bioinformatics Collaborative (NEBC) of the North East Cyberinfrastructure Consortium (NECC), composed of the bioinformatics core facilities from Delaware, Maine, New Hampshire, Rhode Island, and Vermont funded by National Institutes of Health (NIH) Institutional Development Awards (IDeA) and/or National Science Foundation (NSF) Experimental Program to Stimulate Competitive Research (EPSCoR) programs.\n\n\nExisting resources\n\nThere is a single order of holocephalans and 13 orders of elasmobranchs. The distribution of species in orders, families and genera is shown in Figure 1. The batoids are composed of 4 orders, Rajiiformes, Myliobatiformes, Torpidiformes, and Rhinopristiformes, and contain 54% of extant chondrichthyan species. Sharks are broadly divided into two super orders, Galeomorphii and Squalomorphii that together account for 43% of extant chondrichthyan species. The galean sharks include 4 orders: Heterdontiformes, Orectolobiformes, Lamniormes and Carcharhiniformes, and represent 30% of extant chondrichthyan species. Squalean sharks are composed of 4 orders: Squaliformes, Squatiniformes, Pristophoriformes, and Hexanchiformes, comprising 13% of extant chondrichthyan species. Among individual orders, Rajiiformes, the skates, have the most species (345) followed by Carcharhiniformes, the ground sharks (283) and Myliobatiformes (226)2. These ‘big three’ orders contain 854 species, 72% of extant chondrichthyans.\n\nThere is a single order of Holocephalans, Chimaeriformes, and 13 orders of elasmobranchs. The distribution of chondrichthyan species in each of the 14 orders is shown relative to the total number of species, genera and families for the clade. The batoids are composed of 4 orders, Rajiiformes, Myliobatiformes, Torpidiformes, and Rhinopristiformes, and contain 54% of extant chondrichthyan species. Sharks are broadly divided into two super orders, Galeomorphii and Squalomorphii that together include the remaining 9 orders and 43% of extant chondrichthyan species.\n\nChondrichthyan conservation, management, and research all benefit from easily accessible and well-documented molecular resources. The organization of data and metadata in archival databases is critically important for efficient use of large and complex datasets. The International Nucleotide Sequence Database Collaboration (INSDC) is composed of three large public nucleotide repositories, DNA Data Bank of Japan (DDBJ), European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL-EBI), and GenBank at the National Center for Biotechnology Information (NCBI). Recently, two new NCBI database projects were initiated to collect details of samples, BioSample, and project data, BioProject, and propagate the metadata to all associated database entries in an effort to expand the use of already existing and rapidly expanding molecular resources42. Figure 2 illustrates the relationship between BioProject, BioSample and the sequence data for SkateBase. Because the BioProject and BioSample databases were established in 2012, not all existing datasets have metadata or details of the biological source to populate a BioSample and BioProject entry. When available, BioProject and BioSample hyperlinks are included for Sequence Read Archive (SRA), Expressed Sequence Tag (EST) and Genome Survey Sequence (GSS) datasets in the tables below.\n\nA. The little skate genome project is represented as a BioProject entry that connects all samples and data thematically. A BioSample record describes the DNA sample that was used for genome sequencing that was generated from a single stage 32 skate embryo. The SRA catalogs the unassembled Illumina genome sequence data. The Whole Genome Shotgun (WGS) database contains the contiguous sequences from shotgun sequencing projects. The assembled and annotated mitochondrial genome was deposited in GenBank and subsequently included in the NCBI Reference Sequence Database (RefSeq). B. The project to characterize the embryonic transcriptomes of L. erinacea, C. milii and S. canicula is represented in a BioProject entry. Three BioSample entries, one for each species, lead to three SRA datasets. The transcriptome data is represented also in the Gene Expression Omnibus (GEO), a database of high-throughput functional genomic data derived from microarrays and next-generation sequencing technologies.\n\nTable 1 is a summary of chondrichthyan sequence data in NCBI databases, UniProtKB, and the Protein Data Bank (PDB) with L. erinacea, Callorhinchus milii and Scyliorhinus canicula, the three species featured at SkateBase listed individually. The distribution of holocephalans and elasmobranchs in public databases is illustrated in Figure 3. Despite the majority of species belonging to Elasmobranchii, the GenBank, UniProtKB/TrEMBL, and Gene databases are dominated by chimaera data derived from the genome sequence of the elephant shark, C. milii43. Elasmobranch data predominates in UniProtKB/Swiss-Prot, PDB, BioProject and BioSample databases as well as the number of whole mitochondrial genomes (WMG) in GenBank. The EST and SRA databases are nearly equally split between the two subclasses.\n\n(WMG) whole mitochondrial genome, (EST) Expressed Sequence Tags, (lib) libraries (GSS) Genome Survey Sequences, (GEO) Gene Expression Omnibus, (WGS) Whole Genome Shotgun, (SRA) Sequence Read Archive, (WMG) whole mitochondrial genomes, (PDB) Protein Data Bank, * includes 16 unidentified fin entries\n\n1 NCBI databases accessed July 25, 2014, 2 Release 2014_07 of 09-Jul-2014, 3 GEO sample accessions\n\nThe distribution of data for Holocephalii (chimaeras) and elasmobranchii (sharks and rays) subclasses of chondrichthyan fishes does not always reflect their species distribution. The number of species represented in GenBank is representative of the actual species distribution but the amount of data in GenBank is not. Holocephalan data forms the majority of the NCBI Gene, GenBank, Genome Survey Sequence (GSS) and UniProt TrEMBL databases. The number of Short Reach Archive (SRA) experiments and EST sequences in nearly equal for each subclass and the remaining databases are primarily populated by elasmobranch data.\n\n\nChondrichthyan genomes\n\nCurrently there are multiple efforts to sequence an elasmobranch genome in various stages of completion (Table 2); however, only the skate genome project currently has data publically available. Efforts to sequence the whale shark are underway at the Georgia Aquarium and Emory University (personal communication, Alistair Dove, Georgia Aquarium). Genoscope leads a project to sequence the genome of another oviparous elasmobranch, the catshark, S. canicula. The current assembly is described in Table 2. A second version of the catshark genome with 200x coverage, including mate pair sequencing, is in progress (personal communication, Sylvie Mazan, French National Centre for Scientific Research). Among holocephalans, the genome of the elephant shark, C. milii, was first described in a 1.4x coverage assembly in 200644. With continued sequencing the assembly coverage is currently 19.25x and data has been made available through the project website (http://esharkgenome.imcb.a-star.edu.sg/) and Genbank43.\n\n1 (M) Mega or (G) Giga base pairs; (PE) paired end; (est) estimated; (ICMB) Institute of Molecular and Cell Biology, ASTAR, (NECC) North East Cyberinfrastructure Consortium\n\n replaced original sequence data GenBank AAVX00000000.1 (1.4x coverage) released 20-DEC-2006\n\nA powerful resource for characterizing genomes is large-insert clone libraries where each clone contains a large (~100kb) genomic region. Bacterial artificial chromosome (BAC) and P1-derived artificial chromosome (PAC) libraries are DNA constructs within a plasmid used to transform bacteria. As the bacteria grow the inserted DNA is amplified and subsequently isolated and sequenced. BACs are beneficial for genome sequencing projects because the insert size can be very large, nearly 350 kb, facilitating assembly post-sequencing. BAC/PAC libraries were built for several chondrichthyan species including the nurse shark, Ginglymostoma cirratum45; elephant shark, C. milli46; little skate, L. erinacea47; horn shark, Heterdontus francisci48; dogfish shark, Squalus acanthias49,50; and catshark, S. canicula51. These libraries were used to successfully characterize a handful of genomic regions such as little skate HoxA cluster47,52, horn shark Hox A and D clusters53, catshark HoxA, B and D clusters51,52, C. milii HoxA-D clusters54, immunoglobulin receptor IgW C regions30 and neurohypophysial gene loci46.\n\n\nRNA databases\n\nTranscriptome sequencing seeks to characterize all genes expressed in a tissue or set of tissues in a sample. Technologies to identify the complete RNA transcript sequence have developed from studies of a small number of transcripts to comprehensive characterizations. The application of large-scale cDNA cloning of Expressed Sequence Tags (ESTs) gave initial characterizations of 5-prime and/or 3-prime ends of transcripts in several elasmobranchs including L. erinacea and S. acanthias (Table 3). EST sequence data are available in the EST divisions of the GenBank, EMBL and DDBJ databases that make up the INSDC. cDNA clones and their sequences from these EST projects have enabled the complete characterization of the full-length cDNA sequence of several genes. In the last five years, high-throughput RNA sequencing (RNA-Seq) has been applied to comprehensively examine the complete sequence of transcripts in tissues of cartilaginous fishes. Among the most valuable RNA-Seq datasets are those from whole late-stage embryos following organogenesis. Our project has generated these datasets for L. erinacea, S. canicula and C. milii52. Public RNA-Seq data sets can be found in the NCBI Gene Expression Omnibus and Short Read Archive (SRA) databases or the EBI ArrayExpress and European Nucleic Acid (ENA) archives (Table 3 and Table 4).\n\n1 (ICMB) Institute of Molecular and Cell Biology, ASTAR, (HGMP-RC) Human Genome Mapping Project Resource Centre, Hinxton, (Tgen) Translational Genomics Research Institute AZ, USA, (CNRS) National Center for Scientific Research, France, (MDIBL) Mount Desert Island Biological Laboratory, (CPU) China Pharmaceutical University, (MVC) Madras Veterinary College, TANUVAS, (BGI) Beijing Genomics Institute (SSH) Suppressive subtractive hybridization; (mixeda) liver, kidney, brain, testis, ovary, gill, heart, spleen, rectal gland; (mixedb) rectal gland, kidney, brain, testis, ovary, gill, intestine, heart, spleen\n\n genomic data; 1 (SE) single end or (PE) paired end; 2 (M) Mega or (G) Giga base pairs\n\n3 (MDIBL) Mount Desert Island Biological Laboratory, (ICMB) Institute of Molecular and Cell Biology, ASTAR, (LSTM) Liverpool School of Tropical Medicine, (NECC) North East Cyberinfrastructure Consortium\n\n\nMitochondrial genomes\n\nIndividual mitochondrial genes such as cytochrome c oxidase subunit I (CO1 or COX1) and NADH-ubiquinone oxidoreductase chain 2 (NADH2 or MT-ND2) have been used extensively to construct molecular phylogenies55–57. The Fish barcode of life (FISH-BOL) a working group of the International Barcode of Life Project (iBOL), has CO1 barcodes for 54% of elasmobranchs and 62% of holocephalans (http://www.fishbol.org, accessed July 24, 2014). Recently, whole mitochondrial sequences are increasingly popular for their increased granularity when resolving branches of phylogenetic trees1. Whole mitochondrial genome sequences currently are available for 72 species of sharks, skates, rays and chimaeras. These sequences are accessible in the GenBank, EMBL and DDBJ databases summarized in Table 558.\n\nMetazoan Mitochondrial Genomes Accessible dataset Metamiga (http://amiga.cbmeg.unicamp.br/)\n\n\nChondrichthyan Tree of Life\n\nCurrently, molecular data for cartilaginous fishes is being collected as part of the Chondrichthyan Tree of Life project (http://sharksrays.org). The project website currently includes 5 elements: 1) an interactive phylogenetic tree55; 2) scientific illustrations of specimens; 3) range information for all extant species; 4) interactive comparative anatomy through segmented CT scan data; and 5) DNA sequence for 1265 single copy orthologous genes59. Project data will be available in public databases as well as through the project website once collection and analysis is complete (personal communication, Gavin Naylor, Medical University of South Carolina).\n\n\nProtein databases\n\nGiven the improved technologies to characterize full-length transcripts using RNA-Seq, there are increasingly more protein sequence data for chondrichthyans. The UniProt Consortium, consists of groups from the European Bioinformatics Institute (EBI), the Swiss Institute of Bioinformatics (SIB) and the Protein Information Resource (PIR). The consortium maintains the UniProt Knowledgebase (UniProtKB), a comprehensive and standardized catalogue of protein sequences and functional annotation knowledgebase60. Proteins with UniProtKB accessions are first automatically annotated, unreviewed UniProtKB/TrEMBL entries that progress to UniProtKB/Swiss-Prot entries following curator review. Among Chondrichthyes, there are 12 UniProtKB/Swiss-Prot and 20,170 UniProtKB/TrEMBL entries for holocephalans and 264 UniProtKB/Swiss-Prot and 6,299 UniProtKB/TrEMBL entries for elasmobranchs in Release 2014_07 of 09-Jul-2014 of the knowledgebase (Table 1). An unidentified fin sample accounts for 16 UniProtKB/TrEMBL entries that are not included in either Holocephali or Elasmobranchii. PDB, an archive of protein macromolecular structural data, has 178 entries for Chondrichthyes, all elasmobranchs61. Of these, 76% are derived from 2 species from a single family, Torpediniformes, the electric rays, and in total only 10 species are represented in PDB.\n\nThe distribution of data in NCBI databases, PDB, and UniProtKB for chondrichthyan orders is shown in Figure 4. When order Chimaeriformes is included (Figure 4A) the distributions are disproportionate due to the large volume of annotated sequence data from the elephant shark genome. The distributions are repeated exclusively for elasmobranchs. To understand if the data distribution is representative of the number of species in each order, a species distribution is included in each chart. A cladogram (Figure 4B) is linked to the chart legend and illustrates the phylogeny between chondrichthyan orders.\n\nA. The 14 orders of chondrichthyan fish and their relative distribution in public nucleotide and protein databases for Chondrichthyes and Elasmobranchii are shown individually. The species distribution for each Order and GenBank are similar indicating sequence data has been collected for a broad range of chondrichthyans. For Chondrichthyes, the elephant shark genome project data contributes the majority of the data in NCBI Gene, GenBank, Genome Survey Sequence (GSS), and the Short Reach Archive (SRA) databases. The NCBI GSS, GSS libraries, and Protein Data Bank (PDB) are the least diverse with representation of 1–6 of the 14 Orders. The color of each Order as represented in the bar chart is included in the cladogram key with left to right in the bar chart corresponding with top to bottom in the cladogram. B. A cladogram of Chondrichthyes illustrates the phylogeny relationship between the 14 Orders. The color code associated with each Order appears consecutively in the bar chart.\n\n\nSkateBase\n\nSkateBase (http://skatebase.org) is the public portal for the little skate genome project and is a valuable collection of data and learning resources. The NEBC little skate genome project team hosted three week-long workshops and a mitochondrial genome annotation jamboree with the goal of using the project data to develop a bioinformatics aware workforce and foster collaborative and distributed big data research. The lecture materials and worked annotation examples are included at SkateBase for educational use (http://skatebase.org/workshops). The project vitae contains an overview and timeline of the genome project effort along with key personnel, project related publications and presentations, the curation team, and citation information for researchers utilizing the resource in their publication. A Gene Table currently represents manually curated genes derived from workshops and curriculum with extensive annotation evidence. The number of gene entries will continue to grow through usage and expansion of the SkateBase educational modules. Plans to update the annotation interface to enable community annotation by domain experts is planned for the future.\n\nSkateBase provides links to web resources with chondrichthyan data including the Chondrichthyan Tree of Life, Elephant Shark Genome Project (http://esharkgenome.imcb.a-star.edu.sg), the first described genome for a chimaera43, and Vertebrate TimeCapsule, (http://transcriptome.cdb.riken.go.jp/vtcap), a project that aims to develop a gene database to represent evolution and development for vertebrates and currently includes transcriptome data for a hagfish (Eptatretus burger), shark (S. torazame) and birchir (Polypterus senegalus)62. SkateBase data is linked locally as well as from NCBI in the Gene Expression Omnibus (GSE26235), GenBank (AESE010000000) and Sequence Read Archive (SRA026856) to ensure convenient and easy access. A link to the American Elasmobranch Society (http://www.elasmo.org), a non-profit organization with the mission of advancing the scientific study of living and fossil sharks, skates, rays, and chimaeras and promoting education, conservation, and wise utilization of natural resources, connects domain scientists to the little skate genome project.\n\nSkateBase data includes embryonic transcriptomes for three chondrichthyan species, a chimaera, C. milii, a shark, S. canicula and the little skate, L. erinacea as well as the first draft of the little skate genome. The assembled skate genome sequence gave a single high-coverage contiguous sequence that represented the entire length of the mitochondrial genome. The mitochondrial genome was subsequently annotated as part of a Jamboree in 201163. The annotated sequence is represented by the NCBI Reference Sequence (RefSeq) project, accession NC_016429, and provides extensive information for each gene.\n\nWhole embryos were used to build the transcriptome libraries available at SkateBase35. Two C. milii embryos, stage 32, were combined and used to build a chimaera library. The transcriptome library for S. canicula was assembled from six pooled embryos, stages 24–30. The embryonic skate transcriptome library was assembled using six pooled embryos ranging in stage from 20–29. This combination of stages encompasses a large portion of the developmental period for these fishes and represents a catalog of genes important for organogenesis of all or part of every physiological system. Early developmental events are similar for nearly all elasmobranchs regardless of reproductive mode or adult body form enabling the data to be useful for more than just the specific species from where it was derived64. Since all three embryonic transcriptomes contain a similar stage embryo direct comparison for temporal expression patterns is possible. Skatebase includes tools for data investigation, SkateBLAST, a sequence retrieval tool, Skate Contig Lookup, and genome browsers for three skate whole mitochondrial sequences, L. erinacea, the thorny skate, Amblyraja radiata, and, the ocellate spot skate, Okamejei kenojei. Skatebase contains resources that can be used for teaching and research purposes. As an example, two use cases follow, one for sequence or homology based research and the other for education.\n\n\nSkateBLAST\n\nA common task for researchers is searching for genes of interest in a genome or transcriptome. Knowledge of the gene sequence at the DNA or RNA level is needed for many different studies, including phylogenetic analysis or designing primers for quantitative PCR gene expression studies. Here we describe the major steps necessary to identify relevant sequences for a gene of interest using the BLAST sequence similarity tool at SkateBase. SkateBase features a web interface to BLAST, named SkateBLAST, that builds upon the ViroBLAST package version 2.265, with custom modifications allowing parallel cluster-based execution of queries and enhanced display of results. The overall workflow consists of a) entering a query sequence and selecting the database to search; b) evaluating the alignments returned; c) retrieving the sequence from one of the SkateBLAST databases; and d) checking to make sure that the retrieved sequence aligns best to the query sequence. The following description provides a brief tutorial on the overall workflow while describing tools at SkateBase.\n\nFigure 5 demonstrates the use of SkateBLAST to find expressed sequences for the gene, suppressor of cytokine signaling 6 (SOCS6). SOCS6 is a E3 ubiquitin ligase that interacts with c-KIT to suppress cellular proliferation through its SH2 domain66. The first step to identify SOCS6 in the skate transcriptome begins with entering the protein sequence for human SOCS6 that was obtained from UniProt and searching this sequence against the skate transcriptome using the tblastn program. The next step is to evaluate the alignments to determine which transcriptome sequences best represent SOCS6. When interpreting the pairwise alignments from SkateBlast as in any BLAST tool, it is important to examine: a) alignment statistics; b) alignment coverage; and c) presence of protein domains that you may expect to be conserved. The alignment statistics are reported to ascertain whether you would expect the given alignment by chance or not. There are three key alignment statistics, the expectation (E)-value, percent identity and alignment length. The E-value represents the probability that you would expect an alignment with that alignment score or better by random chance, thus the lower the E-value, the better the alignment. Conversely, the greater the percent identity (percent identical sequence) and alignment length, the more similar the two sequences are assumed to be. Alignment coverage with respect to the query or subject sequence (alignment length divided by the length of the query or subject sequence) can also be an important consideration, as low coverage suggests that important regions of one or both sequences may not be represented in the alignment. Finally, there may be particular sequence features, such as protein domains, that you would expect to find in the alignment. If those domains are missing, then it suggests that you have a partial or misleading alignment.\n\nA. SkateBLAST query form showing the four steps to align the UniProt sequence for human SOCS6 (O14544) against the skate embryonic transcriptome using tblastn. Step 1 is to enter the sequence in FASTA format. The second step is to choose the tblastn program that will align the query protein sequence against translated sequences in all six possible reading frames. The third step is to select the embryonic transcriptome as the sequence database to search. The fourth step is to launch the search. B. The complete BLAST output can be accessed by clicking the “Inspect BLAST output” link at the top of the summary report page. This is necessary to examine the sequence alignments. C. Four important fields in the output should be examined carefully to interpret the alignments and determine which returned alignment best represents the skate ortholog to SOCS6. First, the alignment score, E-value, alignment length and percent identity can be used to interpret the overall alignment significance. Alignment coverage with respect to the query protein sequence and the subject transcriptome sequence can be interpreted by comparing the alignment coordinates to the length of the query protein sequence and length of the transcriptome sequence. In this example, the entire query protein sequence is covered by this transcriptome sequence. D. The SkateBase Contig Lookup tool can be used to retrieve the transcriptome sequence found in the SOCS6 tblastn search in FASTA format. Sequences from the skate genome assembly or the skate, S. canicula or C. milii transcriptome assemblies can be retrieved using this tool. E. Output from the NCBI ORF Finder tool showing a 536aa ORF in the skate transcriptome contig that best represents SOCS6 (left). Alignment from blastx search of the skate transcriptome sequence (contig 15542) against human UniProt using NCBI BLAST to validate that the contig aligned best to human SOSC6 rather than another human gene.\n\nOnce a transcriptome sequence of interest, such as contig15542, is identified in the SkateBLAST results, you must do a reciprocal search of that sequence against a database of protein sequences to confirm that the sequence aligns best to your gene of interest. You can retrieve the full sequence directly from the BLAST tool or using the Skate Contig Lookup tool (Figure 5D): a) specify the transcriptome that you had originally searched using SkateBLAST; b) enter the sequence identification or contig number is entered into the query box; and c) select the ‘GO’ button. The user can copy the returned sequence and use it for further exploration of sequence homology at NCBI or similar databases.\n\n\nSkateBase classroom use case: teach concepts of gene and protein annotation\n\nSkateBase includes valuable teaching resources derived from the project workshops on gene and protein annotation. Infrastructure for sequence annotation was developed and modules for use in teaching are available. Access to the teaching modules is through the Curator Access link from the homepage and permission is granted by request using the email link at the bottom of each page, info@SkateBase.org. Once successfully logged into the site, access to pre-computed blast results, guides and examples, annotation forms, and links to external tools helpful for sequence analysis are available. Gene annotation begins with a transcriptome contig identified through a SkateBlast search as illustrated above. The portion of the transcript that codes for protein is identified using an open reading frame or ORF finder tool. Annotation follows a workflow where complimentary sequences from the transcriptome and genome are aligned allowing annotation of both sequences using Sequence Ontology vocabulary67. The evidence is recorded in an annotation form that records information about the annotator and sequences and includes a comment box for questions and comments between students and teachers or curators and annotators. The annotation form records the pairwise alignment of the transcriptome and genome contigs, notes concerning mismatches or gaps, as well as output from the ORF tool. The untranslated regions (UTR) at the beginning and end of each sequence, 5’UTR and 3’ UTR regions, as well as the intron/exon structure for the genomic contig and CDS for the transcriptomic contig are recorded in the Gene Annotation Form. When completing the Gene Annotation Form, the appropriate activity must be selected and can be customized to specify the user’s course ID, institution or workshop title to track annotation history. Protein annotation uses the rapid annotation interface for proteins, RACE-P, developed by the PIR. A UniProt accession number is required to initiate a new annotation form. The form is composed of 6 blocks of information, protein information, gene information, a bibliography, Gene Ontology (GO), computational analysis using tools such as Pfam68, TMHMM69, SignalP70, COILS71, NetPhos72 and EMBOSS73, and protein family evidence.\n\n\nDiscussion\n\nThe volume of data in GenBank continues to grow exponentially, doubling nearly every 18 months. The first sequences for chondrichthyes appeared in 1983 and the overall data trend for chondrichthyans is similar to all of GenBank with three exceptions. First, the rate of increase is less than GenBank. Second, the number of sequences deposited during the first decade of the 21st century was nearly stagnant in comparison. Third, a large spike is observed in late 2012 attributed to the Elephant Shark Genome Project data (Figure 6). Molecular data is increasingly important for all aspects of research utilizing chondrichthyan fishes74. It can be a forensic tool to understand species when fins are landed without carcass and ensure protected species and quotas are respected75–79. For migrating species molecular data serves as a surrogate to classical tagging data to understand population structure and range80–83. In studies of evolution, molecular data provides estimates of divergence time and supplements morphological and ecological traits as the basis for a phylogeny. The benefits and uses of molecular data for these fishes are limited only by the amount of data available. SkateBase provides the only genomic data publically available for an elasmobranch in addition to embryonic transcriptomes, data tools, and educational resources.\n\nGenBank is the National Institutes of Health (NIH) genetic sequence database and together with the DNA Databank of Japan (DDBJ) and the European Molecular Biological Laboratory (EMBL) comprise the International Nucleotide Sequence Database Collaboration (INSDC). The cumulative base pair total for all taxa as well as chondrichthyan only data are given versus time for GenBank and Whole Genome Shotgun (WGS) data. The Elephant Shark Genome Project is responsible for the spike in chondrichthyan GenBank in 2011. The little skate and elephant shark genome projects are currently the only two WGS datasets (yellow line).\n\nSequencing projects require significant funding and personnel commitments but generate a large amount of information that can be translated to knowledge by domain experts. The efficiency of this process is affected most by allowing the scientific community to access the data. The value of data sharing can be measured by the number of publications that result from its distribution. To date, 19 publications in peer-reviewed journals have used data derived from SkateBase (http://skatebase.org/vitae). Molecular data are the means to investigate genes and develop reagents for gene expression studies by PCR or in situ hybridization. Small scale sequencing efforts that generate limited or fragmented data often get deposited to hard disks and remain ‘buried’ and out of reach. Efforts to deposit this data at public sequence repositories are encouraged to build the foundation of data required to describe this dynamic and ancient clade of fishes. We invite investigators to contact the authors in an effort to survey the volume of private data available for potential distribution through SkateBase.\n\nThe transcriptome data at SkateBase serves as a platform to teach molecular techniques, technologies, and bioinformatics in the context of studying elasmobranchs. As next generation sequencing (NGS) technologies evolve it is important for scientists and students to understand how the sequence was generated and caveats of workflow for each data type in order to recognize errors and customize analysis algorithms. The educational materials and infrastructure at SkateBase have been used by University of Delaware, Georgetown University, MDI Biological Laboratory, University of Maine at Machias, University of Rhode Island, and most recently the Virginia Institute of Marine Science to teach gene and protein annotation concepts. We invite and look forward to continued expansion of the SkateBase educational platform as we refine the infrastructure and expand the data available for investigation through continued sequencing efforts.", "appendix": "Author contributions\n\n\n\nThe little skate genome project is a collaborative effort with participation from all authors. JTW and BLK prepared the first draft of the manuscript. SWP oversaw the design and implementation of the SkateBase web portal and SkateBLAST tool. All authors were involved with the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nThe authors declare to have no competing interests.\n\n\nGrant information\n\nThis work was supported by a re-entry career award to JTW, National Institute of General Medical Sciences (NIGMS) IDeA Networks of Biomedical Research Excellence (INBRE) 3P20GM103446-12S1. Skate genome sequencing was funded by National Institutes of Health (NIH) National Center for Research Resources (NCRR) ARRA Supplements to 5P20RR016463-12 (MDIBL), 5P20RR016472-12 (UD), and 5P20RR16462 (UVM). The North East Cyberinfrastructure Consortium (NECC) is funded by NIH NCRR grants 5P20RR016463-12 (MDIBL), 5P20RR016472-12 (UD), 5 P20 RR16462 (UVM), 5P20RR016457-11 (URI), and 5P20RR030360-03 (UNH) and NIH NIGMS grants 8 P20 GM103423-12 (MDIBL), 8P20GM103446-12 (UD), 8P20GM103449 (UVM), 8 P20 GM103430-11 (URI), and 8P20GM103506-03 (Dartmouth), and NSF Experimental Program to Stimulate Competitive Research (EPSCoR) grants EPS-0904155 (UM), EPS-081425 (UD), EPS-1101317 (UVM), EPS-1004057 (URI), and EPS-1101245 (UNH).\n\n\nAcknowledgements\n\nThe authors thank the North East Bioinformatics Collaborative of the North East Cyberinfrastructure Consortium and Karl Steiner and Steven Stanhope of DE-INBRE, Patricia Hand of ME-INBRE, and Judith VanHouten of VT-INBRE. We thank the Delaware Biotechnology Institute (DBI) for hosting the website, database and NECC Shared Data Center. We thank Karol Miaskiewicz for systems administration support and Gang Li and Zhiwen Li for web development. We thank Bruce Kingham and the DBI Sequencing and Genotyping Center. Sylvie Mazan, Alistair Dove, Dave Ebert, Cecelia Arighi, Qinghua Wang, James Sulikowski and Eric Haenni contributed data, samples, artwork or editorial comments. We thank Gavin Naylor and the Chondrichthyan Tree of Life project team including Lindsay Marshall (illustrations), Jason Davies (database, computational work, and visualizations), Will White and Peter Last (maps and taxonomy), Shannon Corrigan and Lei Yang (gene capture data), and Callie Crawford and Thomas Fussell (CT scanning and anatomy) for providing a description of the project scope.\n\n\nReferences\n\nInoue JG, Miya M, Lam K, et al.: Evolutionary origin and phylogeny of the modern holocephalans (Chondrichthyes: Chimaeriformes): a mitogenomic perspective. Mol Biol Evol. 2010; 27(11): 2576–86. PubMed Abstract \| Publisher Full Text\n\nEbert DA, Ho H, White WT, et al.: Introduction to the systematics and biodiversity of sharks, rays, and chimaeras (Chondrichthyes) of Taiwan. 2013; 3752(1): 5–19. Publisher Full Text\n\nStevens J, Bonfil R, Dulvy NK, et al.: The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J Mar Sci. 2000; 57(3): 476–94. Publisher Full Text\n\nGarcía VB, Lucifora LO, Myers RA: The importance of habitat and life history to extinction risk in sharks, skates, rays and chimaeras. Proc Biol Sci. 2008; 275(1630): 83–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nField IC, Meekan MG, Buckworth RC, et al.: Chapter 4. Susceptibility of sharks, rays and chimaeras to global extinction. Adv Mar Biol. 2009; 56: 275–363. PubMed Abstract \| Publisher Full Text\n\nSpringer S: Oviphagous Embryos of the Sand Shark, Carcharias taurus. Copeia. 1948; 1948(3): 153. Reference Source\n\nJoung SJ, Chen CT, Clark E, et al.: The whale shark, Rhincodon typus, is a livebearer: 300 embryos found in one ‘megamamma’ supreme. Environ Biol Fishes. 1996; 46(3): 219–23. Publisher Full Text\n\nStevens J, Bonfil R, Dulvy NK, et al.: The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J Mar Sci. 2000; 57(3): 476–94. Publisher Full Text\n\nLiu SY, Chan CL, Lin O, et al.: DNA barcoding of shark meats identify species composition and CITES-listed species from the markets in Taiwan. PLoS One. 2013; 8(11): e79373. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDudgeon CL, Blower DC, Broderick D, et al.: A review of the application of molecular genetics for fisheries management and conservation of sharks and rays. J Fish Biol. 2012; 80(5): 1789–843. PubMed Abstract \| Publisher Full Text\n\nPinhal D, Shivji MS, Nachtigall PG, et al.: A streamlined DNA tool for global identification of heavily exploited coastal shark species (genus Rhizoprionodon). PLoS One. 2012; 7(4): e34797. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarbuto M, Galimberti A, Ferri E, et al.: DNA barcoding reveals fraudulent substitutions in shark seafood products: The Italian case of “palombo” (Mustelus spp.). Food Res Int. 2010; 43(1): 376–81. Publisher Full Text\n\nFeldheim KA, Gruber SH, Dibattista JD, et al.: Two decades of genetic profiling yields first evidence of natal philopatry and long-term fidelity to parturition sites in sharks. Mol Ecol. 2014; 23(1): 110–7. PubMed Abstract \| Publisher Full Text\n\nSchneider I, Aneas I, Gehrke AR, et al.: Appendage expression driven by the Hoxd Global Control Region is an ancient gnathostome feature. Proc Natl Acad Sci U S A. 2011; 108(31): 12782–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGillis JA, Dahn RD, Shubin NH: Shared developmental mechanisms pattern the vertebrate gill arch and paired fin skeletons. Proc Natl Acad Sci U S A. 2009; 106(14): 5720–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGillis JA, Modrell MS, Baker CVH: Developmental evidence for serial homology of the vertebrate jaw and gill arch skeleton. Nat Commun. 2013; ; 4: 1436. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHentschel DM, Bonventre JV: Novel non-rodent models of kidney disease. Curr Mol Med. 2005; 5(5): 537–46. PubMed Abstract \| Publisher Full Text\n\nSteele SL, Yancey PH, Wright PA: Dogmas and controversies in the handling of nitrogenous wastes: osmoregulation during early embryonic development in the marine little skate Raja erinacea; response to changes in external salinity. J Exp Biol. 2004; 207(Pt 12): 2021–31. PubMed Abstract \| Publisher Full Text\n\nStolte H, Galaske RG, Eisenbach GM, et al.: Renal tubule ion transport and collecting duct function in the elasmobranch little skate, Raja erinacea. J Exp Zool. 1977; 199(3): 403–10. PubMed Abstract \| Publisher Full Text\n\nEvans DH: A brief history of the study of fish osmoregulation: the central role of the Mt. Desert Island Biological Laboratory. Front Physiol. 2010; 1: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWyffels JT, Luer CA, Walsh CJ, et al.: In vivo exposure of clearnose skates, Raja eglanteria, to ionising X-radiation: acute effects on the peripheral blood, spleen, and epigonal and Leydig organs. Fish Shellfish Immunol. 2007; 23(2): 401–18. PubMed Abstract \| Publisher Full Text\n\nWyffels JT, Walsh CJ, Luer CA, et al.: In vivo exposure of clearnose skates, Raja eglanteria, to ionizing X-radiation: acute effects on the thymus. Dev Comp Immunol. 2005; 29(4): 315–31. PubMed Abstract \| Publisher Full Text\n\nLutton BV, Callard IP: Morphological relationships and leukocyte influence on steroid production in the epigonal organ-ovary complex of the skate, Leucoraja erinacea. J Morphol. 2008; 269(5): 620–9. PubMed Abstract \| Publisher Full Text\n\nWalsh CJ, Luer CA, Bodine AB, et al.: Elasmobranch immune cells as a source of novel tumor cell inhibitors: Implications for public health. Integr Comp Biol. 2006; 46(6): 1072–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEason DD, Litman RT, Luer CA, et al.: Expression of individual immunoglobulin genes occurs in an unusual system consisting of multiple independent loci. Eur J Immunol. 2004; 34(9): 2551–8. PubMed Abstract \| Publisher Full Text\n\nAnderson MK, Pant R, Miracle AL, et al.: Evolutionary origins of lymphocytes: ensembles of T cell and B cell transcriptional regulators in a cartilaginous fish. J Immunol. 2004; 172(10): 5851–60. PubMed Abstract \| Publisher Full Text\n\nCai SY, Soroka CJ, Ballatori N, et al.: Molecular characterization of a multidrug resistance-associated protein, Mrp2, from the little skate. Am J Physiol Regul Integr Comp Physiol. 2003; 284(1): R125–30. PubMed Abstract\n\nClusin W, Spray DC, Bennett MVL: Activation of a voltage-insensitive conductance by inward calcium current. Nature. 1975; 256(5516): 425–7. PubMed Abstract \| Publisher Full Text\n\nElger M, Hentschel H, Litteral J, et al.: Nephrogenesis is induced by partial nephrectomy in the elasmobranch Leucoraja erinacea. J Am Soc Nephrol. 2003; 14(6): 1506–18. PubMed Abstract \| Publisher Full Text\n\nZhang C, Du Pasquier L, Hsu E: Shark IgW C region diversification through RNA processing and isotype switching. J Immunol. 2013; 191(6): 3410–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeard J: The origin and histogenesis of the thymus in Raja batis. Zool Jahrb Abt Anat Ontog Tiere. 1902; 17: 403–80. Reference Source\n\nLuer CA, Walsh CJ, Bodine AB, et al.: The elasmobranch thymus: Anatomical, histological, and preliminary functional characterization. J Exp Zool. 1995; 273(4): 342–54. Publisher Full Text\n\nWyffels JT, Walsh CJ, Luer CA, et al.: In vivo exposure of clearnose skates, Raja eglanteria, to ionizing X-radiation: acute effects on the thymus. Dev Comp Immunol. 2005; 29(4): 315–31. PubMed Abstract \| Publisher Full Text\n\nFeldheim KA, Chapman DD, Sweet D, et al.: Shark virgin birth produces multiple, viable offspring. J Hered. 2010; 101(3): 374–7. PubMed Abstract \| Publisher Full Text\n\nChapman DD, Shivji MS, Louis E, et al.: Virgin birth in a hammerhead shark. Biol Lett. 2007; 3(4): 425–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson DP, Baverstock W, Al-Jaru A, et al.: Annually recurring parthenogenesis in a zebra shark Stegostoma fasciatum. J Fish Biol. 2011; 79(5): 1376–82. PubMed Abstract \| Publisher Full Text\n\nCompagno LJV: Alternative life-history styles of cartilaginous fishes in time and space. Environ Biol Fishes. 1990; 28(1–4): 33–75. Publisher Full Text\n\nLuer CA, Gilbert PW: Mating behavior, egg deposition, incubation period, and hatching in the clearnose skate, Raja eglanteria. Environ Biol Fishes. 1985; 13(3): 161–71. Publisher Full Text\n\nLuer CA, Walsh CJ, Bodine AB, et al.: Normal embryonic development in the clearnose skate, Raja eglanteria, with experimental observations on artificial insemination. Environ Biol Fishes. 2007; 80(2–3): 239–55. Publisher Full Text\n\nMasuda M, Izawa Y, Kametuta S, et al.: Artificial insemination of the cloudy catshark. J Japanese Assoc Zool Gard Aquariums. 2003; 44(2): 39–43. Reference Source\n\nCastro JI, Bubucis PM, Overstrom NA, et al.: The Reproductive Biology of the Chain Dogfish, Scyliorhinus retifer. Copeia. 1988; 1988(3): 740. Publisher Full Text\n\nBarrett T, Clark K, Gevorgyan R, et al.: BioProject and BioSample databases at NCBI: facilitating capture and organization of metadata. Nucleic Acids Res. 2012; 40(Database issue): D57–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVenkatesh B, Lee AP, Ravi V, et al.: Elephant shark genome provides unique insights into gnathostome evolution. Nature. 2014; 505(7482): 174–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVenkatesh B, Kirkness EF, Loh YH, et al.: Survey sequencing and comparative analysis of the elephant shark (Callorhinchus milii) genome. PLoS Biol. 2007; 5(4): e101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLuo M, Kim H, Kudrna D, et al.: Construction of a nurse shark (Ginglymostoma cirratum) bacterial artificial chromosome (BAC) library and a preliminary genome survey. BMC Genomics. 2006; 7: 106. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGwee PC, Tay BH, Brenner S, et al.: Characterization of the neurohypophysial hormone gene loci in elephant shark and the Japanese lamprey: origin of the vertebrate neurohypophysial hormone genes. BMC Evol Biol. 2009; 9: 47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMulley JF, Zhong YF, Holland PW: Comparative genomics of chondrichthyan Hoxa clusters. BMC Evol Biol. 2009; 9: 218. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChiu C, Amemiya C, Dewar K, et al.: Molecular evolution of the HoxA cluster in the three major gnathostome lineages. Proc Natl Acad Sci U S A. 2002; 99(8): 5492–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMattingly C, Parton A, Dowell L, et al.: Cell and molecular biology of marine elasmobranchs: Squalus acanthias and Raja erinacea. Zebrafish. 2004; 1(2): 111–20. PubMed Abstract \| Publisher Full Text\n\nMerson RR, Mattingly C, Planchart AJ: Tandem duplication of aryl hydrocarbon receptor (AHR) genes in the genome of the spiny dogfish shark (Squalus acanthias). Bull MDIBL. 2009; 48: 43–4. Reference Source\n\nOulion S, Debiais-Thibaud M, d’Aubenton-Carafa Y, et al.: Evolution of Hox gene clusters in gnathostomes: insights from a survey of a shark (Scyliorhinus canicula) transcriptome. Mol Biol Evol. 2010; 27(12): 2829–38. PubMed Abstract \| Publisher Full Text\n\nKing BL, Gillis JA, Carlisle HR, et al.: A natural deletion of the HoxC cluster in elasmobranch fishes. Science. 2011; 334(6062): 1517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim CB, Amemiya C, Bailey W, et al.: Hox cluster genomics in the horn shark, Heterodontus francisci. Proc Natl Acad Sci U S A. 2000; 97(4): 1655–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRavi V, Lam K, Tay BH, et al.: Elephant shark (Callorhinchus milii) provides insights into the evolution of Hox gene clusters in gnathostomes. Proc Natl Acad Sci U S A. 2009; 106(38): 16327–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNaylor GJP, Caira JN, Jensen K, et al.: Elasmobranch Phylogeny: A mitochondrial estimate based on 595 species. In: Carrier JC, Musick JA, Heithaus MR, editors. The Biology of Sharks and Their Relatives. CRC Press, Taylor & Francis Group. 2012; 31–56. Reference Source\n\nNaylor GJP, Ryburn JA, Fedrigo O, et al.: Phylogenetic relationships among the Major Lineages of Modern Elasmobranchs. 1995. Reference Source\n\nNaylor GJP, Caira JN, Jensen K, et al.: A DNA Sequence–Based Approach To the Identification of Shark and Ray Species and Its Implications for Global Elasmobranch Diversity and Parasitology. Bull Am Museum Nat Hist. 2012; 367: 1–262. Publisher Full Text\n\nFeijão PC, Neiva LS, de Azeredo-Espin AM, et al.: AMiGA: the arthropodan mitochondrial genomes accessible database. Bioinformatics. 2006; 22(7): 902–3. PubMed Abstract \| Publisher Full Text\n\nLi C, Hofreiter M, Straube N, et al.: Capturing protein-coding genes across highly divergent species. Biotechniques. 2013; 54(6): 321–6. PubMed Abstract\n\nUniProt Consortium. Activities at the Universal Protein Resource (UniProt). Nucleic Acids Res. 2014; 42(Database issue): D191–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerman HM, Westbrook J, Feng Z, et al.: The Protein Data Bank. Nucleic Acids Res. 2000; 28(1): 235–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakechi M, Takeuchi M, Ota KG, et al.: Overview of the transcriptome profiles identified in hagfish, shark, and bichir: current issues arising from some nonmodel vertebrate taxa. J Exp Zool B Mol Dev Evol. 2011; 316(7): 526–46. PubMed Abstract \| Publisher Full Text\n\nWang Q, Arighi CN, King BL, et al.: Community annotation and bioinformatics workforce development in concert--Little Skate Genome Annotation Workshops and Jamborees. Database (Oxford). 2012; 2012: bar064. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJennifer TW: Embryonic development of Chondrichthyan fishes – a review. In: W KY, Lue CA & Kapoor BG, editors. Development of Non-Teleost Fishes. Enfield: Science Publishers. 2009; 1–103. Publisher Full Text\n\nDeng W, Nickle DC, Learn GH, et al.: ViroBLAST: a stand-alone BLAST web server for flexible queries of multiple databases and user’s datasets. Bioinformatics. 2007; 23(17): 2334–6. PubMed Abstract \| Publisher Full Text\n\nZadjali F, Pike AC, Vesterlund M, et al.: Structural basis for c-KIT inhibition by the suppressor of cytokine signaling 6 (SOCS6) ubiquitin ligase. J Biol Chem. 2011; 286(1): 480–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMungall CJ, Batchelor C, Eilbeck K: Evolution of the Sequence Ontology terms and relationships. J Biomed Inform. 2011; 44(1): 87–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFinn RD, Bateman A, Clements J, et al.: Pfam: the protein families database. Nucleic Acids Res. 2014; 42(Database issue): D222–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrogh A, Larsson B, von Heijne G, et al.: Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol. 2001; 305(3): 567–80. PubMed Abstract \| Publisher Full Text\n\nPetersen TN, Brunak S, von Heijne G, et al.: SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods. 2011; 8(10): 785–6. PubMed Abstract \| Publisher Full Text\n\nLupas A, Van Dyke M, Stock J: Predicting coiled coils from protein sequences. Science. 1991; 252(5009): 1162–4. PubMed Abstract \| Publisher Full Text\n\nBlom N, Gammeltoft S, Brunak S: Sequence and structure-based prediction of eukaryotic protein phosphorylation sites. J Mol Biol. 1999; 294(5): 1351–62. PubMed Abstract \| Publisher Full Text\n\nRice P, Longden I, Bleasby A: EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet. 2000; 16(6): 276–7. PubMed Abstract \| Publisher Full Text\n\nHeist E: Genetics of Sharks, Skates, and Rays. In: Musick J, Carrier J, Heithaus M, editors. Biology of Sharks and Their Relatives. CRC Press. 2004; 471–85. Publisher Full Text\n\nChapman D, Pinhal D, Shivji M: Tracking the fin trade: genetic stock identification in western Atlantic scalloped hammerhead sharks Sphyrna lewini. Endanger Species Res. 2009; 9: 221–8. Publisher Full Text\n\nClarke SC, Magnussen JE, Abercrombie DL, et al.: Identification of shark species composition and proportion in the Hong Kong shark fin market based on molecular genetics and trade records. Conserv Biol. 2006; 20(1): 201–11. PubMed Abstract \| Publisher Full Text\n\nLiu SY, Chan CL, Lin O, et al.: DNA barcoding of shark meats identify species composition and CITES-listed species from the markets in Taiwan. PLoS One. 2013; 8(11): e79373. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShivji MS: DNA Forensic Applications in Shark Management and Conservation. In: Carrier JC., Musick JA., Heithaus MR., editors. Sharks and Their Relatives II Biodiversity, Adaptive Physiology, and Conservation. CRC Press. 2010; 593–610. Publisher Full Text\n\nRodrigues-Filho LF, Pinhal D, Sodré D, et al.: Applications and Impacts on Genetic Diversity. In: Caliskan M, editor. Analysis of Genetic Variation in Animals. InTech. 2012; 269–86. Publisher Full Text\n\nChabot CL, Allen LG: Global population structure of the tope (Galeorhinus galeus) inferred by mitochondrial control region sequence data. Mol Ecol. 2009; 18(3): 545–52. PubMed Abstract \| Publisher Full Text\n\nKeeney DB, Heist EJ: Worldwide phylogeography of the blacktip shark (Carcharhinus limbatus) inferred from mitochondrial DNA reveals isolation of western Atlantic populations coupled with recent Pacific dispersal. Mol Ecol. 2006; 15(12): 3669–79. PubMed Abstract \| Publisher Full Text\n\nSchultz JK, Feldheim KA, Gruber SH, et al.: Global phylogeography and seascape genetics of the lemon sharks (genus Negaprion). Mol Ecol. 2008; 17(24): 5336–48. PubMed Abstract \| Publisher Full Text\n\nWard RD, Holmes BH, White WT, et al.: DNA barcoding Australasian chondrichthyans: results and potential uses in conservation. Mar Freshw Res. 2008; 59(1): 57–71. Publisher Full Text" }	[ { "id": "5793", "date": "26 Aug 2014", "name": "Kevin Feldheim", "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 publicly accessible database SkateBase, a public site for the little skate genome project. This paper is well written, and I have just a few comments and suggestions: Why was the little skate picked for genome size? Is it small and manageable? Is it a typical size compared to other chondrichthyans? Page 3, 1st column: change “compliance with fishing quotas” to “maintain compliance with fishing quotas”. Page 3, 2nd column: Heterodontiformes is misspelled (Heterdontiformes). Page 3, 2nd column: Lamniformes is misspelled (Lamniormes). Page 3, under “Existing resources” I suggest including Chimaeriformes here as well. Page 7: Heterodontus is misspelled (Heterdontus). Page 8: species name for hagfish should be burgeri (not burger). Page 8: The last three sentences under the “Protein databases” section can be incorporated into the legend for Figure 4. Page 8: bichir is misspelled (birchir). Page 8: change “SkateBase data is linked” to “SkateBase data are linked” and “SkateBase data includes” to “SkateBase data include”. Page 19, 1st column: change “this data” to “these data”. Figures and Tables: In the figure 1 legend — most of this information is stated in the text. I suggest keeping the sentence: “The distribution of chondrichthyan species in each of the 14 orders is shown relative to the total number of species, genera and families for the clade.” I also suggest explaining the figure in a little more detail. What do the numbers mean? Do the colors represent anything? Figure 2 legend: “is” should not be italicized. Table 1: What do the numbers in the columns mean? Are these numbers of entries? Table 2: Contigs for L. erinacea is 2,62,265. Also, under N50, change 1466 to 1,466. Figure 4: It is difficult to tell apart colors for the following pairs:Squatiniformes and CarcharhiniformesLamniformes and MyliobatiformesChimaeriformes and Rajiiformes This can be rectified by using black, white and yellow for three of the above orders in the figure. Figure 5 legend: I suggest changing “Four important fields in the output should be examined carefully to interpret the alignments and determine which returned alignment best represents the skate ortholog to SOCS6. First, the alignment score, E-value, alignment length and percent identity can be used to interpret the overall alignment significance.” to “Four important fields in the output should be examined carefully to interpret the alignments: the alignment score, E-value, alignment length and percent identity.”", "responses": [] }, { "id": "5798", "date": "01 Sep 2014", "name": "Ed Heist", "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 has two components: a description of current genomic, proteomic, and transcriptomic resources for chondrichthyan fishes and a description of the SkateBase project. Of the two, I would rather have seen more emphasis on SkateBase since the rate at which new data are being generated will make the survey data presented in the manuscript increasingly incomplete. Nevertheless I think it is a good paper worthy of publication and SkateBase seems to be a valuable resource.Specific comments include:Abstract – capitalize Sarcopterygii, Actinopterygii misspelled. Throughout the manuscript there are numerous inconsistencies in capitalization, e.g. “order Chimaeriformes” and “Fish barcode of life (FISH-BOL) a working group of the International Barcode of Life Project (iBOL)” both on page 8.Introduction – In the first sentence replace “Chondrichthyan fishes” with “Living chondrichthyan fishes.”Paragraph 2 – The text implies that the only benefits of chondrichthyans are ecotourism and fisheries. Perhaps a mention of ecological services (e.g. food-web dynamics) or other benefits of chondrichthyan conservation would be appropriate here.Paragraph 4 – “asexual” parthenogenesis seems redundant.Paragraph 5 – perhaps some more details on the distribution, life history, reproductive mode, genome size etc. of little skate is in order. Why is the geographic location of the species relevant?Figure 1 is hard for me to decipher. I can’t figure out what the two separate outer rings indicate and why the spacing between them is different for the orders vs. the families, genera, and species. I really think this figure is too complex and tries to cram too much information that is peripheral to the goals of the paper into a single figure. I may be old fashioned, but I think a table that listed the numbers of families, genera, and species would have been sufficient. There also needs to be a citation on the source of the information as the taxonomy and number of recognized species of chondrichthyans is changing. Figure 4B – A citation is needed for the cladogram.", "responses": [] } ]	1	https://f1000research.com/articles/3-191
https://f1000research.com/articles/3-188/v1	12 Aug 14	{ "type": "Data Note", "title": "A revised method for measuring distraction by tactile stimulation", "authors": [ "Jacqueline R. Schechter", "Deanna J. Greene", "Jonathan M. Koller", "Kevin J. Black", "Jacqueline R. Schechter", "Deanna J. Greene", "Jonathan M. Koller" ], "abstract": "Sensory hypersensitivity (SH) refers to the tendency to attend to subtle stimuli, to persist in attending to them, and to find them noxious. SH is relatively common in several developmental disorders including Tourette Syndrome and Chronic Tic Disorder (TS/CTD). This study was an attempt to quantify the extent to which a mild tactile stimulus distracts one’s attention in TS/CTD. Fourteen adults with TS/CTD and 14 tic-free control subjects completed questionnaires regarding SH and ADHD, and TS/CTD subjects completed self-report measures of current and past tic disorder symptoms and of current obsessions and compulsions. All subjects performed a sustained attention choice reaction time task during alternating blocks in which a mildly annoying stimulus (von Frey hair) was applied to the ankle (“ON”) or was not applied (“OFF”). We present here the clinical and cognitive task data for each subject.", "keywords": [ "Tourette Syndrome and Chronic Tic Disorder (TS/CTD) are complex neuropsychiatric disorders in which patients present with multiple motor or vocal tics. In many patients with TS/CTD", "attentional problems are present and ADHD is the most common condition that co-occurs with TS/CTD", "occurring in about 50% of patients (Greimel et al.", "2011). These attentional problems can range from mild to severe and can impact the patient’s ability to complete tasks", "sustain attention", "and keep track of personal items. Sensory hypersensitivity (SH", "the tendency to be sensitive to subtle stimuli that most people would no longer attend to after habitation has occurred) is also commonly present in patients with TS/CTD (Belluscio et al.", "2011). Patients with sensory hypersensitivity have difficulty tuning out otherwise neutral stimuli such as the tag in their shirt or the voices of people around them. In addition", "patients with TS/CTD often experience a premonitory urge before a tic which some patients describe as a feeling of itchiness", "pressure", "tenseness", "or energy. These tics and premonitory urges can also be distracting (Kane", "1994)." ], "content": "Introduction\n\nTourette Syndrome and Chronic Tic Disorder (TS/CTD) are complex neuropsychiatric disorders in which patients present with multiple motor or vocal tics. In many patients with TS/CTD, attentional problems are present and ADHD is the most common condition that co-occurs with TS/CTD, occurring in about 50% of patients (Greimel et al., 2011). These attentional problems can range from mild to severe and can impact the patient’s ability to complete tasks, sustain attention, and keep track of personal items. Sensory hypersensitivity (SH; the tendency to be sensitive to subtle stimuli that most people would no longer attend to after habitation has occurred) is also commonly present in patients with TS/CTD (Belluscio et al., 2011). Patients with sensory hypersensitivity have difficulty tuning out otherwise neutral stimuli such as the tag in their shirt or the voices of people around them. In addition, patients with TS/CTD often experience a premonitory urge before a tic which some patients describe as a feeling of itchiness, pressure, tenseness, or energy. These tics and premonitory urges can also be distracting (Kane, 1994).\n\nThe observation that SH can distract one from a cognitively demanding task suggested the approach we used in a small pilot study to attempt to quantify SH by examining the effects of a tactile stimulus on reaction time during a sustained attention task (Panagopoulos et al., 2013). The present study was designed to improve on some of the earlier methods and to compare subjects with TS/CTD to tic-free controls, with the goal of quantifying the extent to which sensory hypersensitivity in patients with TS/CTD affects sustained attention. To do this, participants with and without TS/CTD performed an attention task in the presence and absence of a subtle sensory stimulus.\n\n\nMethods\n\nThis study was approved by the Washington University Human Research Protection Office (IRB), proposal # 201108081.\n\nA convenience sample of 14 adults with TS/CTD and 14 tic-free adults participated in the study. Subjects completed a questionnaire that included age, sex, and the question, “Have you ever been diagnosed with any of the following,” with check boxes for ADHD (Attention-Deficit/Hyperactivity Disorder), learning disorder, atopic (allergic) dermatitis, OCD (Obsessive-Compulsive Disorder), Tourette syndrome, other tic disorder, and other neurological illness (to be specified by the respondent). All participants also completed the 26-item Adult Sensory Questionnaire (ASQ), developed to screen for sensory defensiveness in adults (Kinnealey et al., 1995), and the ADHD Rating Scale (Barkley, 1998; Magnusson et al., 2006). Subjects endorsing tics also rated symptom severity for the past week using the Yale Global Tic Severity Scale (YGTSS; Leckman et al., 1989), the Yale-Brown Obsessive Compulsive Scale (Goodman et al., 1989a,b), and the Premonitory Urge for Tics Scale (PUTS; Woods et al., 2005). These data were collected using REDCap electronic data capture tools hosted at Washington University (Harris et al., 2009). For tic subjects, the Diagnostic Confidence Index (DCI) was also completed; the DCI assesses typical historical features of TS (Robertson et al., 1999).\n\nThe subjects were seated at a fixed distance in front of a laptop computer in a darkened room. All subjects then performed a 12-minute choice reaction time task consisting of pressing one key on a computer keyboard when the capital letter ‘S’ appeared on the screen and pressing another key when the numeral 5 appeared on the screen. They were instructed to respond as quickly and accurately as possible with the right hand. The task consisted of 11 blocks; the first was a 2-minute OFF condition and the remaining blocks were each 1 minute long alternating between the ON and the OFF condition. Throughout the ON condition, a 4.74N von Frey hair was held against a point previously marked on the subject’s ankle, with ~1 Hz mild increases of pressure to just bend the von Frey hair. During the OFF condition, the von Frey hair was absent. E-Prime® 2 software was used to present stimuli and to collect all task data (www.pstnet.com/eprime.cfm; Schneider et al., 2002a,b).\n\n\nData availability\n\nF1000Research: Dataset 1. A revised method for measuring distraction by tactile stimulation: Subject characteristics and task data, 10.5256/f1000research.4944.d34156 (Schechter et al., 2014).", "appendix": "Author contributions\n\n\n\nAll authors contributed to conception of the study and design of the experiments. JRS and DJG carried out the research. JRS and KJB prepared the first draft 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 project was supported by National Institutes of Health (NIH) grants K24 MH087913, K01 MH104592, P30 CA091842, UL1 TR000448, the Siteman Comprehensive Cancer Center, and a Tourette Syndrome Association fellowship (DJG). The content is solely the responsibility of the authors and does not necessarily represent the official view of the funders.\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\nBarkley RA: Attention deficit/hyperactivity disorder: A handbook for diagnosis and treatment. 2nd edition. New York; Guilford Press; 1998. Reference Source\n\nBelluscio BA, Jin L, Watters V, et al.: Sensory sensitivity to external stimuli in Tourette syndrome patients. Mov Disord. 2011; 26(14): 2538–2543. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris PA, Taylor R, Thielke R, et al.: Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2): 377–381. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoodman WK, Price LH, Rasmussen SA, et al.: The Yale-Brown Obsessive Compulsive Scale. I. Development, use and reliability. Arch Gen Psychiatry. 1989a; 46(11): 1006–1011.PubMed Abstract \| Publisher Full Text\n\nGoodman WK, Price LH, Rasmussen SA, et al.: The Yale-Brown Obsessive Compulsive Scale. II. Validity. Arch Gen Psychiatry. 1989b; 46(11): 1012–1016. PubMed Abstract \| Publisher Full Text\n\nGreimel E, Wanderer S, Rothenberger A, et al.: Attentional performance in children and adolescents with tic disorder and co-occurring attention-deficit/hyperactivity disorder: new insights from a 2 × 2 factorial design study. J Abnorm Child Psychol. 2011; 39(6): 819–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKane MJ: Premonitory urges as “attentional tics” in Tourette’s syndrome. J Am Acad Child Adolesc Psychiatry. 1994; 33(6): 805–808. PubMed Abstract \| Publisher Full Text\n\nKinnealey M, Oliver B, Wilbarger P: A phenomenological study of sensory defensiveness in adults. Am J Occup Ther. 1995; 49(5): 444–451. PubMed Abstract \| Publisher Full Text\n\nLeckman JF, Riddle MA, Hardin MT, et al.: The Yale Global Tic Severity Scale: initial testing of a clinician-rated scale of tic severity. J Am Acad Child Adolesc Psychiatry. 1989; 28(4): 566–573. PubMed Abstract \| Publisher Full Text\n\nMagnusson P, Smari J, Sigurdardottir D, et al.: Validity of self-report and informant rating scales of adult ADHD symptoms in comparison with a semistructured diagnostic interview. J Atten Disord. 2006; 9(3): 494–503. PubMed Abstract \| Publisher Full Text\n\nPanagopoulos VN, Greene DJ, Campbell MC, et al.: Towards objectively quantifying sensory hypersensitivity: a pilot study of the “Ariana effect”. PeerJ. 2013; 1: e121. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobertson MM, Banerjee S, Kurlan R, et al.: The Tourette syndrome diagnostic confidence index: development and clinical associations. Neurology. 1999; 53(9): 2108–2112. PubMed Abstract \| Publisher Full Text\n\nSchechter JR, Greene DJ, Koller JM, et al.: Dataset 1. A revised method for measuring distraction by tactile stimulation: Subject characteristics and task data. F1000Research. 2014. Data Source\n\nSchneider W, Eschman A, Zuccolotto A: E-Prime Reference Guide. Pittsburgh: Psychology Software Tools Inc. 2002a. Reference Source\n\nSchneider W, Eschman A, Zuccolotto A: E-Prime User’s Guide. Pittsburgh: Psychology Software Tools Inc. 2002b. Reference Source\n\nWoods DW, Piacentini J, Himle MB, et al.: Premonitory Urge for Tics Scale (PUTS): Initial psychometric results and examination of the premonitory urge phenomenon in youths with Tic disorders. J Devel Behav Pediatr. 2005; 26(6): 397–403. PubMed Abstract \| Publisher Full Text" }	[ { "id": "6060", "date": "08 Sep 2014", "name": "Andrea Cavanna", "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 and original research project. The methodology is sound and the quality of the data is high.", "responses": [ { "c_id": "968", "date": "08 Sep 2014", "name": "Kevin J Black", "role": "Author Response", "response": "We appreciate Prof. Cavanna's assessment." } ] }, { "id": "6165", "date": "18 Sep 2014", "name": "Euripedes Constantino Miguel", "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 and relevant study that examines the effects of a tactile stimulus on reaction time during a sustained attention task in patients with Tourette syndrome/chronic tic disorder (TS/CTD) vs. non-TS/CTD subjects in order to quantify the extent to which sensory hypersensitivity influences attentional outcome.Sensory hypersensitivity has a conceptual overlap with what we call sensory phenomena 1-4. We define sensory phenomena by subjective uncomfortable or unpleasant sensations, feelings, and/or perceptions that may precede or accompany repetitive behaviors not only in TS but also in some common comorbid conditions such as obsessive-compulsive disorder (OCD). Types of sensory phenomena include:Physical sensations (tactile and/or muscle-joint), defined as uncomfortable sensations localized in a specific region of the body (skin, muscles, or joints), which precede or occur along with repetitive behaviors; “Just right” perceptions, triggered by tactile, visual, or auditory sensations, including the need for things to feel, look or sound “just right”; Feelings of incompleteness (i.e., inner feeling or perception of discomfort that makes the patient do things indefinitely, or until feeling ‘complete’); Energy release (i.e., generalized inner tension or energy that builds up and needs to be released by an action) and Urge only (i.e., “have to do it” perception related to the repetitive behavior).In the last 2 decades, our group has developed a brief instrument to assess the presence and the severity of sensory phenomena (the University of São Paulo Sensory Phenomena Scale) to investigate OCD phenotypic subtypes and its relationship with TS/CTD. This instrument has been validated in English 5. OCD patients with sensory phenomena have more comorbidity with tics, early-onset at OC symptoms, more symptoms of ordering and symmetry and low insight 6,7. Though the sample size of the Schechter et al. study is not large, it will be very interesting to examine the role of OCD (YBOCS Scores) as a confounder in attentional outcomes, as well as whether sensory hypersensitivity mediates the expression or severity of OCD symptoms. Finally, it would be interesting to observe how sensory hypersensitivity predicts TS treatment response and its comorbid conditions, as well as how specific treatments to these disorders influence sensory hypersensitivity.Therefore, we believe that the investigation of subjective experiences such as sensory hypersensitivity is relevant to better refine the phenomenological presentation of TS/CTD patients, allowing the identification of more homogeneous subgroups and the mechanisms involved in treatment response. This understanding may help to reveal etiological factors as well as the development of more efficacious therapeutic interventions for a spectrum of patients that share common psychopathology involving sensorial experiences.\n\nWe are eager to read the final results of the present work. Acknowledgement:Both Drs Miguel and Hoexter receive a grant from the Brazilian Agencies for Research: FAPESP and CNPq.", "responses": [ { "c_id": "996", "date": "18 Sep 2014", "name": "Kevin J Black", "role": "Author Response", "response": "We thank Drs. Miguel and Hoexter for their thoughtful, forward-looking commentary, and appreciate their suggestions on phenomenology and characterizing comorbidity. I have followed earlier reports on the USP-SPS with great interest, but was unaware of the recent publication on the English version. It would undoubtedly help place sensory hypersensitivity in better context with other sensory phenomena that occur commonly in OCD and tic disorders, and will be of interest to our future work along this line." } ] } ]	1	https://f1000research.com/articles/3-188
https://f1000research.com/articles/3-130/v1	19 Jun 14	{ "type": "Research Article", "title": "Ecological implications of reduced pollen deposition in alpine plants: a case study using a dominant cushion plant species", "authors": [ "Anya Reid", "Robyn Hooper", "Olivia Molenda", "Christopher J. Lortie", "Anya Reid", "Robyn Hooper", "Olivia Molenda" ], "abstract": "The reproductive assurance hypothesis states that self-incompatible female plants must produce twice the number of seeds relative to their self-compatible hermaphroditic counterparts to persist in gynodioecious populations. This is a viable life-history strategy, provided that pollination rates are sufficiently high. However, reduced pollination rates in alpine plants are likely due to climate induced plant-pollinator mismatches and general declines in pollinators. Using a gynodioecious population of the dominant plant Silene acaulis (Caryophyllaceae), we tested the reproductive assurance hypothesis and also the stress gradient hypothesis with a series of pollinator exclusion trials and extensive measurements of subsequent reproductive output (gender ratio, plant size, percent fruit-set, fruit weight, seeds per fruit, total seeds, seed weight, and seed germination). The reproductive assurance hypothesis was supported with female plants being more sensitive to and less likely to be viable under reductions in pollination rates. These findings are the first to show that the stress gradient hypothesis is also supported under a gradient of pollen supply instead of environmental limitations. Beneficiary abundance was negatively correlated to percent fruit-set under current pollen supply, but became positive under reduced pollen supply suggesting that there are important plant-plant-pollinator interactions related to reproduction in these alpine plant species.", "keywords": [ "alpine", "cushion plant", "pollination", "climate induced mismatch", "pollinator decline" ], "content": "Introduction\n\nAt least two future climate change scenarios predict that pollination rates will be directly reduced, and these are due to either pollinator declines (Potts et al., 2010) and/or plant-pollinator mismatches (Hegland et al., 2009). Recently there has been concern over general global trends of reduced pollinator species abundance and diversity that are both predicted to reduce pollination rates to plants (Memmott et al., 2007; Potts et al., 2010). Climate induced plant-pollinator mismatch can reduce pollination rates by creating a temporal mismatch in pollinator emergence and plant flowering times (Hegland et al., 2009). Pollinator emergence is regulated by temperature, whereas plant bloom time is regulated by photoperiod (Hegland et al., 2009). If climate warming shifts pollinator emergence but not plant bloom time, then a temporal mismatch between plants and pollinators occurs (Hegland et al., 2009). This scenario is likely more pronounced in alpine and polar environments that are experiencing a more rapid increase in annual temperature than the global average (very high confidence; Intergovernmental Panel on Climate Change [IPCC], 2013). Conceivably, both of these reductions in pollination rates occur simultaneously and thus adaptability of different sexual morphs in alpine plants can be an important consideration in predicting responsiveness and variation in reproductive output.\n\nHypotheses associated with pollen availability in alpine environments are controversial. It has been assumed that pollination rates are inherently low in alpine environments (Larson & Barrett, 2000; Totland & Sottocornola, 2001; Torres-Díaz et al., 2011). This is attributed to low temperatures, overcast conditions, strong winds, and relatively unpredictable weather being challenging for insect pollinators (Körner, 1999). These harsh conditions generally lead to lower pollinator diversity, abundance, and activity in alpine ecosystems relative to milder ecosystems (Kevan, 1972; Primack, 1978; Moldenke & Lincoln, 1979; Arroyo et al., 1982; Primack, 1983; Billings, 1987; Totland, 1993). Alternatively, pollination rates can increase with elevations, suggesting adequate pollen availability under current conditions (Arroyo et al., 1985; Arroyo & Squeo, 1990; Bingham & Orthner, 1998; Utelli & Roy, 2000).\n\nThe reproductive assurance hypothesis (RAH) and the stress gradient hypothesis (SGH) are thus highly relevant hypotheses to explore in better understanding climate change effects on alpine communities. The RAH proposes that when pollen supply is low, self-compatible plants are favored over self-incompatible plants (Lloyd, 1992; Lloyd & Schoen, 1992). This is because self-compatible plants create their own pollen thereby being more adapted to low or variable pollination rates (Larson & Barrett, 2000; Muñoz & Arroyo, 2006; García-Camacho & Totland, 2009). Further, it has been proposed that self-compatible plants are less likely to become extinct if pollinators drastically decrease or disappear from a given system (Richards, 1997; Morgan et al., 2005). Therefore, self-compatible plants may become favored in the future if pollination supply declines. The SGH is also an important ecological theory to consider with respect to potential climate impacts on pollinators. The SGH states that facilitation between plant species is more common when resources are limited (Bertness & Callaway, 1994). Typically, the SGH is tested using environmental limitations such as temperature or moisture in the alpine (He et al., 2013; Liczner & Lortie, 2014; McIntire & Fajardo, 2014), but has not been applied to the concept of pollen supply as an important limitation for plants in stressful environments. Taken together, these ecological theories provide a solid platform to build pollen limitation studies upon and also provide a set of potential ecological drivers that can help better predict pollination rate changes in the alpine.\n\nHere, we use a gynodioecious population of Silene acaulis to assess the sensitivity of different genders to pollen limitation. We test the following predictions associated with the reproductive assurance hypothesis: (1) that the reproduction of self-incompatible plants is more sensitive to reduced pollen deposition than self-compatible plants and (2) that self-incompatible plants will be less viable under experimentally reduced pollen loads compared to self-compatible plants. In doing so, we also explore whether the SGH applies to the plant-pollinator system in the alpine. Specifically, we predict that facilitation between plants is more common under reduced pollen, i.e. that less pollen can be a novel stressor for alpine plants and that this can in turn relate to plant-plant interactions.\n\n\nMaterials and methods\n\nS. acaulis (L.) Jacq. (Caryophyllacae), commonly known as moss campion, is a common long-lived evergreen cushion that is found throughout the northern hemisphere (Hitchcock & Maguire, 1947). Each plant has a single strong taproot, and there is no clonal reproduction (Morris & Doak, 1998). Small pink flowers can be abundant. S. acaulis is visited by bumblebees (Shykoff, 1988; Shykoff, 1992; Marr, 1997; Delph et al., 1999; Delph & Carroll, 2001), moths, beetles, ants (Marr, 1997; Delph & Carroll, 2001), flies (Totland, 1993; Delph & Carroll, 2001), butterflies and Osmia bees (Reid & Lortie, 2012). S. acaulis is a nurse plant species that, like many other cushion forming plants, benefits other plant species (called beneficiaries) by reducing abiotic stress (Arroyo et al., 2003; Bertness & Callaway, 1994; Callaway & Walker, 1997; Cavieres et al., 2006). Recent beneficiary removal studies suggest that by facilitating the beneficiaries, cushions bear a cost in reduced reproductive success (Cranston et al., 2012; Schöb et al., 2014).\n\nS. acaulis is sexually polymorphic (Hitchcock & Maguire, 1947), and the population examined here was gynodioecious with plants that only have hermaphrodite flowers and other that only have female flowers. Female flowers have three styles with stigmatic lobes and hermaphrodite flowers have ten stamens. Male-sterility in female morphs of S. acaulis is predominantly under nuclear-cytoplasmic control (Delph & Carroll, 2001); the gene for male-sterility is passed on through the female gamete (Lewis, 1941). In addition to S. acaulis, the flowering plant species Antennaria alpina, Arnica sp., Carex sp., Erigeron sp., Luzula sp., Phacelia sericea, Phlox diffusa, Phyllodoce spp., Poa alpina, Potentilla diversifolia, P. heptaphila, P. villosa, Ranunculus eschsoltzii, Saxifraga bronchialis, and Solidago multiradiata were present at relatively high densities.\n\nThe experiments were conducted on the Whistler Mountain in British Columbia 50°03′31.68″N, 122°57′22.53″W, 2168m elevation), Canada, during the snow-free season of July and August 2010. This area is classified as alpine tundra with ten months of snow cover per year (Pojar et al., 1987). A total of 273 S. acaulis plants were measured. Three S. acaulis plants were excluded from the study because they were infected with the pollinator-transmitted anther smut-fungus Microbotryum violaceum that renders the flowers of both genders sterile (Baker, 1947; Alexander & Antonovics, 1988; Hermanutz & Innes, 1994; Marr, 1997).\n\nBefore bud-burst, S. acaulis plants were covered with cloth mesh to prevent pollinators contacting the flowers (Donnelly et al., 1998). As the plant gender was unknown when initially covering, 60 plants were covered to ensure that there would be sufficient replicate plants of each gender. Plant gender was established after bud-burst, and at that time, plants were randomly assigned a treatment such as covered with mesh or open to insect pollinators and marked with a unique identification code. Reduced pollination treatments were the ones covered with mesh and were applied to 20 hermaphrodite and 20 female plants. The first 20 female and male plants found were used as replicates with the additional 20 plants being uncovered. The 40 plants (20 of each gender) selected for the reduced pollination treatments remained covered with mesh for the entire flowering season to exclude all insect pollination.\n\nAll flowers of the reduced-pollination treatment plants were hand-pollinated with pollen collected from S. acaulis plants within 10 meters from the treatment plants. All hermaphrodite flowers with mature anthers were collected in the morning of the hand pollination days. Pollen was then applied using small paintbrushes or by directly touching the anthers to the stigmas of all the treatment-plant flowers. We found that direct contact of the anthers to the stigmas was the most effective method of hand pollination. The exact amount of pollen applied to each flower at each hand-pollination event was not quantified. Hand pollination was repeated on three different days between July 20th and August 1st, 2010.\n\nReproductive output measures were collected from the 40 hand-pollinated treatment plants as well as 231 naturally pollinated S. acaulis plants. These measures included total number of flowers, total number of fruits, percent fruit-set, seeds per fruit, fruit weight, and seed weight. The percentage of germination was calculated on a subset of 60 plants, including the 40 treatment plants and 20 naturally pollinated plants.\n\nThe total number of flowers was counted during fruit collection including both successfully and unsuccessfully (i.e. no fruit) pollinated flowers. Fruits were collected when mature but not yet dehiscing, so that the seeds remained in the fruit capsule. This occurred between August 11th and 25th. All fruits were placed in small labeled paper envelopes and were allowed to dry at room temperature to avoid decomposition. The percentage of fruit-set was calculated using the measures of total number of fruits and total number of flowers. The mean fruit weight (g) was calculated by averaging the weight of five randomly selected fruits per plant. These five fruits were dissected and the seed counted. The mean seed number per fruit was calculated from the seed counts. The total seed number per plant was estimated by multiplying total number of fruits with mean seed number per fruit. Mean seed weight (mg) was calculated by averaging the weight of ten randomly selected seeds per plant. All weighing was done to four significant digits. When a plant produced less than five fruits or ten seeds, the average was based on the maximum number of fruit or seed produced. Weighed seed was stored separately and cold-stratified at 4°C for two months. A test germination trial was conducted with limited success likely because the cold stratification was not sufficient. Therefore, seeds were then stored at 0°C for two additional months in preparation for germination trials.\n\nGermination trials were conducted on the weighed and cold stratified seeds from the 40 S. acaulis plants in reduced pollination treatments and the remaining 20 labeled plants that were left open to natural pollination. Growth chambers were set to standard optimum growing conditions of 20°C and light for 12 hours, then 10°C and dark for the remaining 12 hours of the day (Baskin & Baskin, 1998). Relative humidity was set to 90%. The ten seeds from an individual plant were placed on a labeled filter paper in a Petri dish. Seeds were checked weekly for three months, after which germination is rare (Milbau et al., 2009). Germination was considered to have occurred when the radical broke open the seed (Milbau et al., 2009). Germinated seeds were removed to speed-up counting during the subsequent weeks and reduce counting errors (Milbau et al., 2009). Percent germination was expressed as the fraction of total number of germinated seeds with respect to the total number of seeds per Petri dish.\n\nCushion area and floral density were measured because of their possible effect on reproductive output. Cushion area was defined by the external boundary of vegetation and calculated as an ellipse with the formula,\n\ncushion area = (a/2)(b/2)π\n\nwhere a is the longest diameter of the plant and b the diameter perpendicular to a. We calculated the floral density by dividing the total flower number by the cushion area.\n\nTo test if facilitation became more common under reduced pollen loads (stress gradient hypothesis), we measured beneficiary abundance on all cushion plants. Beneficiary abundance is the total number of individual plants living on the cushions.\n\nTo assess the sensitivity of female and hermaphrodite reproductive output under reduced pollen deposition, we calculated the percent change of reproductive success measures within each gender between current and reduced pollen deposition. Percent change was calculated using the following equation:\n\npercent change = (T-C)/C100\n\nwhere T is the reproductive output measures under the reduced pollination treatments and C is the reproductive output measures under the current pollination rates (Ayres, 1993). Negative numbers indicate that reduced pollination treatments decrease reproductive success and positive numbers indicate that reduced pollination treatments increase reproductive success. This method facilitates comparisons of the direction and magnitude of change.\n\nTo statistically test if reduced pollination, gender, and their interaction effects significantly explained the variation in the reproductive output measures, we used a generalized linear model (GLM) with Poisson distribution and a log link function. Covariate measures of S. acaulis included surface area, S. acaulis floral density and beneficiary abundance.\n\nHermaphrodites provide one-half of the genetic material to the population through pollen production. Therefore, female plants must produce at least twice the number of seeds as hermaphrodite plants to persist in the population (Charnov, 1982) or have offspring that are more fit (Lewis, 1941). To statistically analyze the viability of females under current and reduced pollen deposition levels, we compared female reproductive output to twice that of hermaphrodite reproductive output. In this way, if female reproduction (F) is greater than two times hermaphrodite reproduction (2H), then females are viable in the population. For females to be viable, not all measures of reproductive success need to be twice that of hermaphrodites, but all measures are shown to be comprehensive. GLMs were also used to test if gender significantly affected reproductive output measures under the current ambient and experimentally reduced pollination regimes.\n\nInstead of testing the effect of beneficiaries on measures of reproductive success over a range of environmental gradients (Bertness & Callaway, 1994) we tested whether variation in the current and experimentally reduced pollen supply can be viewed as potential stressor. To test the stress gradient hypothesis in this plant-pollinator system, we conducted correlation analysis and tested for significance in the interaction term between the effect pollination and beneficiary abundance on percent fruit set in a GLM. A significant p-value (p<0.05) indicates that the response of percent fruit set to beneficiary abundance significantly differs between plants in the current and reduce pollination regimes. All analyses are appropriate for dealing with the unbalanced number of replicates between the current and reduced pollination regimes and were done in JMP 10 (SAS, 2012).\n\n\nResults\n\nPollination regime significantly influenced percent fruit-set, seeds per fruit, fruit weight and percent germination (Table 1). Percent fruit-set, seeds per fruit, fruit weight, and percent germination decreased in both genders with reduced pollen deposition (Figure 1). There was a significant interaction effect between gender and pollination regime for percent fruit set, and percent germination (Table 1) indicating that these measures differed in their response. The direction of these differences was significant and is illustrated in the percent change calculations. Percent fruit-set was more significantly reduced in females relative to hermaphrodites (Figure 1), whereas percent germination reduced to a greater in hermaphrodites compared to females (Figure 1).\n\nSummary of GLM results testing the effect of gender, pollination regime, and the gender by pollination regime interaction on measures of reproductive success with covariate measures of S. acaulis surface area (SA), S. acaulis floral density, and beneficiary abundance indicated by . Significance is considered at p < 0.05 and is indicated in bold.\n\n± 1 standard error bars shown.\n\nUnder current pollination rates, female plants had more than twice (2.98 times) the percent fruit-set compared to hermaphrodites (Figure 2). Female plants had less than twice the seeds/fruit, fruit weight, seed weight and percent germination compared to hermaphrodites (Figure 2). Under current pollination rates, all reproductive output measures were significantly different between females and two times hermaphrodite reproductive measures (Table 2).\n\n± 1 standard error bars shown.\n\nSummary of GLM results testing the effect of gender on measures of reproductive success under the current pollination regime with covariate measures of S. acaulis surface area (SA), S. acaulis floral density, and beneficiary abundance indicated by . Reproductive measures for hermaphrodites are doubled. Significance is considered at p < 0.05 and is indicated in bold.\n\nUnder reduced pollination rates, none of the female reproductive measures were greater than two times that of hermaphrodites (Figure 2). Under reduced pollination rates, all reproductive output measures, except seed weight and percent germination, were significantly different between females and two times hermaphrodite reproductive measures (Table 3).\n\nSummary of GLM results testing the effect of gender on measures of reproductive success under the reduced pollination regime with covariate measures of S. acaulis surface area (SA), S. acaulis floral density, and beneficiary abundance indicated by . Reproductive measures for hermaphrodites are doubled. Significance is considered at p > 0.05 and indicated in bold.\n\nBeneficiary abundance had a significant effect on percent fruit-set (Table 1). Under the current pollination rates, percent fruit-set and beneficiary abundance are negatively related (slope = -0.48, R2 = 0.04, Figure 3). In contrast, under reduced pollination rates, percent fruit-set and beneficiary abundance are positively related (slope = 0.24, R2 = 0.06, Figure 3). The slopes of these lines significantly differ (Chi2 161.25, p-value <0.0001).\n\nThe linear best-fit line for plants under current pollination regime (black dashed line) has a negative slope of -0.48 and R2 value of 0.04. The linear best-fit line for plants under reduced pollination regime (grey line) has a positive slope of 0.24 and R2 value of 0.06. The slopes of these lines significantly differ (Chi2 161.25, p-value <0.0001).\n\n\nDiscussion\n\nPollinator declines and climate mismatches are important in understanding the capacity for alpine plants to respond to possible future scenarios with reduced pollen loads. Using the cushion plant S. acaulis, we tested two predictions associated with the reproductive assurance hypothesis and more broadly we investigated whether pollination stress influences plant-plant interactions. All predictions were supported. Females were more sensitive than hermaphrodites to reduced pollen loads resulting in reproductive output dropping to below twice that of hermaphrodites. As the pollen supply conditions became more stressful (i.e. reduced), beneficiary plant species on these cushions positively related to the percent fruit-set of the cushions. Hence, the reproductive assurance hypothesis and use of pollen reduction experiments can be important tools for ecological experiments on the responsiveness of alpine plant-pollinator systems to future changes in pollen availability. Importantly, loss of keystone alpine plant species such as cushions may in turn have significant and reciprocal negative impacts on the pollinator communities.\n\nPollinator declines and plant-pollinator mismatches are important potential drivers of broad plant-community dynamics in the alpine if dominant cushion plant species are impacted because they often function as keystone plant species (Arroyo et al., 2003; Bertness & Callaway, 1994; Callaway & Walker, 1997; Cavieres et al., 2006; Molenda et al., 2012; Butterfield et al., 2013; McIntire & Fajardo, 2014). Current trends of decreasing native pollinator populations are a pressing concern globally (Memmott et al., 2007; Potts et al., 2010; Bartomeus & Winfree, 2013). In these alpine environments, bumblebees in particular are suggested to be critical because they are the most effective alpine pollinator in these ecosystems (Bingham & Orthner, 1998; Chittka et al., 1999; Gegear & Laverty, 1998), and for instance, because some alpine bumblebee populations are in decline (Colla & Ratti, 2010). The future scenario that pollinators may emerge before flowers are in bloom due to a warming climate has also been proposed (Hegland et al., 2009) and shown in the alpine environments of Japan (Kudo, 2013). Although it may not mimic the exact future pollen deposition rates, the experimental design tested herein begins to explore how alpine plants may respond to reduced-pollen loads. Differences in reproduction between alpine plant genders are thus a critical avenue of research and are important because cushion plants are common facilitators (for instance, see meta-analysis in Liczner & Lortie, 2014).\n\nTwo findings were particularly useful from an ecological perspective. Female cushion plants became less viable under reduced pollen loads and the stress gradient hypothesis was supported as means to model a gradient of pollen deposition rates because it is also clearly a potential limitation in some stressful ecosystems. Percent fruit-set was the only measure that indicated a drop in female viability. Indeed, this more than compensated for female plants’ reproductive disadvantage over hermaphrodite plants. Hence, a reasonable proxy or single measure to consider in similar future studies using dominant cushion plants is percent fruit-set only. Interestingly, our results also supported the application of the stress gradient hypothesis to pollen limitation in addition to its original formulation for environmental stress or consumer pressure. The relationship between beneficiary abundance and percent fruit-set shifted from negative to positive as the pollination rates were reduced. This supports previous findings that under current pollination to the plant community, beneficiary plants living on cushions generally have a cost associated with the cushion plants’ reproductive fitness (Cranston et al., 2012; Schöb et al., 2013). The findings here however also further suggest that under reduced pollen loads this cost of facilitation can be diminished likely because competition between cushions and the other species is significantly reduced. Clearly, additional research is needed to identify the causal relationships between plant-plant interactions and plant-pollinator interactions with dominant plant species that host other species in stressful environments such as the alpine.\n\n\nData availability\n\nF1000Research: Dataset 1. The reproductive effects of reduced pollen deposition via exclosures and hand pollination on the cushion plant Silene acaulis, 10.5256/f1000research.4382.d29313 (Reid et al., 2014).", "appendix": "Author contributions\n\n\n\nAR and CJL wrote the paper and analyzed the data. AR, RH and OM performed the experiments and collected the data. All authors read and approved the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was funded by an NSERC DG and an NSERC Canadian Pollination Initiative (CANPOLIN) grant to CJL. AR was funded by an Ontario Graduate Scholarship and two York University Faculty of Graduate Studies research grants. RH was funded by NSERC URSA. OM was funded through York University Department of Biology.\n\n\nAcknowledgements\n\nThis is publication #26 of NSERC-CANPOLIN.\n\n\nReferences\n\nAlexander HM, Antonovics J: Disease spread and population dynamics of anther-smut infection of Silene alba cause by the fungus Ustilago violacea. J Ecol. 1988; 76: 91–104. Publisher Full Text\n\nArroyo MTK, Armesto JJ, Primack RB: Community studies in pollination ecology in the high temperate Andes of Central Chile II. Effect of temperature on visitation rates on pollination possibilities. Plant Syst Evol. 1985; 149(3–4): 187–203. Publisher Full Text\n\nArroyo MTK, Cavieres LA, Peñaloza A, et al.: Positive associations between the cushion plant Azorella monantha (Apiaceae) and alpine plant species in the Chilean Patagonian Andes. Plant Ecol. 2003; 169(1): 121–129. Publisher Full Text\n\nArroyo MTK, Primack R, Armesto J: Community studies in pollination ecology in the high temperate Andes of Central Chile. I. Pollination mechanisms and altitudinal variation. Amer J Bot. 1982; 69(1): 82–97. Publisher Full Text\n\nArroyo MTK, Squeo F: Relationship between plant breeding systems and pollination. Academic Press, London. 1990. Reference Source\n\nAyres MP: Plant defense, herbivory, and climate change. in Kareiva, P.M., Kingsolver, J.G. and Huey, R.B., editors. Biotic interactions and global change. Sinauer Associates, Sunderland, M.A. 1993; 75–94. Reference Source\n\nBaker HG: Infection of species of Melandrium by Ustilago violacea (Pers) Fuckel and the transmission of the resultant disease. Ann Bot. 1947; 11(3): 333–348. Reference Source\n\nBartomeus I, Winfree R: Pollinator declines: reconciling scales and implications for ecosystem services. F1000Research. 2013; 2: 146. Publisher Full Text\n\nBaskin CC, Baskin JM: Seeds: ecology, biogeography, and evolution of dormancy and germination. Academic Press, London. 1998. Reference Source\n\nBertness M, Callaway RM: Positive interactions in communities. Trends Ecol Evol. 1994; 9(5): 191–193. PubMed Abstract \| Publisher Full Text\n\nBillings WD: Constraints to plant growth, reproduction, and establishment in arctic environments. Arct Alp Res. 1987; 19(4): 357–365. Publisher Full Text\n\nBingham RA, Orthner AR: Efficient pollination of alpine plants. Nature. 1998; 391: 238–239. Publisher Full Text\n\nButterfield BJ, Cavieres LA, Callaway RM, et al.: Alpine cushion plants inhibit the loss of phylogenetic diversity in severe environments. Ecol Lett. 2013; 16(4): 478–486. PubMed Abstract \| Publisher Full Text\n\nCallaway RM, Walker LR: Competition and facilitation: a synthetic approach to interactions in plant communities. Ecology. 1997; 78(7): 1958–1965. Publisher Full Text\n\nCavieres LA, Badano EI: Do facilitative interactions increase species richness at the entire community level? J Ecol 2009; 97(6): 1181–1191. Publisher Full Text\n\nCavieres LA, Badano EI, Sierra-Almeida A, et al.: Positive interactions between alpine plant species and the nurse cushion plant Laretia acaulis do not increase with elevation in the Andes of central Chile. New Phytol. 2006; 169(1): 59–69. PubMed Abstract \| Publisher Full Text\n\nCharnov EL: The theory of sex allocation. Princeton University Press, Princeton, N.J. 1982. Reference Source\n\nChittka L, Thomson JD, Waser NM: Flower constancy, insect psychology, and plant evolution. Naturwissenschaften. 1999; 86(8): 361–377. Publisher Full Text\n\nColla SR, Ratti CM: Evidence for the decline of the western bumble bee (Bombus occidentalis Greene) in British Columbia. Pan-Pacific Entomologist. 2010; 86(2): 32–34. Publisher Full Text\n\nCranston BH, Callaway RM, Monks A, et al.: Gender and abiotic stress affect community-scale intensity of facilitation and its costs. J Ecol. 2012; 100: 915–922. Publisher Full Text\n\nDelph LF, Bailey MF, Marr DL: Seed provisioning in gynodioecious Silene acaulis (Caryophyllaceae). Am J Bot. 1999; 86(1): 140–144. PubMed Abstract \| Publisher Full Text\n\nDelph LF, Carroll SB: Factors affecting relative seed fitness and female frequency in a gynodioecious species, Silene acaulis. Evol Ecol Res. 2001; 3: 487–505. Reference Source\n\nDonnelly SE, Lortie CJ, Aarssen LW: Pollination in Verbascum thapsus (Scrophulariaceae): the advantages of being tall. Am J Bot. 1998; 85(11): 1618–1625. PubMed Abstract\n\nGarcía-Camacho R, Totland Ø: Pollen limitation in the alpine: A meta-analysis. Arct Antarc Alp Res. 2009; 41(1): 103–111. Publisher Full Text\n\nGegear RJ, Laverty TM: How many flower types can bumble bees work at the same time? Can J Zool. 1998; 76(7): 1358–1365. Publisher Full Text\n\nHe Q, Bertness MD, Altieri AH: Global shifts towards positive species interactions with increasing environmental stress. Ecol Lett. 2013; 16(5): 695–706. PubMed Abstract \| Publisher Full Text\n\nHegland SJ, Nielsen A, Lázaro A, et al.: How does climate warming affect plant-pollinator interactions? Ecol Lett. 2009; 12(2): 184–195. PubMed Abstract \| Publisher Full Text\n\nHermanutz LA, Innes DJ: Gender variation in Silene acaulis (Caryophyllaceae). Plant Syst Evol. 1994; 191(1–2): 69–81. Publisher Full Text\n\nHitchcock CL, Maguire B: A revision of the North American species of Silene. Seattle University of Washington Press, Seattle. 1947.\n\nIPCC. 5th Annual Report. Climate change 2013: The physical science basis. Intergovernmental panel on climate change. 2013. Reference Source\n\nKevan PG: Insect pollination of high arctic flowers. J Ecol. 1972; 60(3): 831–847. Publisher Full Text\n\nKörner C: Alpine plant life. 2nd edition. Springer, Berlin. 1999. Publisher Full Text\n\nKudo G: Vulnerability of phenological synchrony between plants and pollinators in an alpine ecosystem. Ecol Res. 2013. Publisher Full Text\n\nLarson BMH, Barrett SCH: A comparative analysis of pollen limitation in flowering plants. Biol J Linn Soc. 2000; 69(4): 503–520. Publisher Full Text\n\nLewis DG: Male sterility in natural populations of hermaphrodite plants. New Phytologist. 1941; 40(1): 56–63. Publisher Full Text\n\nLiczner AR, Lortie CJ: A global meta-analytic contrast of cushion-plant effects on plants and on arthropods. PeerJ. 2014; 2: e265. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLloyd DG: Self- and cross-fertilization in plants. II. The selection of self-fertilization. Int J Plant Sci. 1992; 153(3): 370–380. Publisher Full Text\n\nLloyd DG, Schoen DJ: Self- and cross-fertilization in plants. I. Functional dimensions. Int J Plant Sci. 1992; 153(3): 358–369. Publisher Full Text\n\nMarr DL: Impact of a pollinator-transmitted disease on reproduction in healthy Silene acaulis. Ecology. 1997; 78(5): 1471–1480. Publisher Full Text\n\nMcIntire EJB, Fajardo A: Facilitation as a ubiquitous driver of biodiversity. New Phytol. 2014; 201(2): 403–416. PubMed Abstract \| Publisher Full Text\n\nMemmott J, Craze PG, Waser NM, et al.: Global warming and the disruption of plant-pollinator interactions. Ecol Lett. 2007; 10(8): 710–717. PubMed Abstract \| Publisher Full Text\n\nMilbau A, Graae BJ, Shevtsova A, et al.: Effects of a warmer climate on seed germination in the subarctic. Ann Bot. 2009; 104(2): 287–296. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoldenke AR, Lincoln PG: Pollination Ecology in Montane Colorado: A Community Analysis. Phytologia. 1979; 42(4): 349–379. Reference Source\n\nMolenda O, Reid A, Lortie CJ: The alpine cushion plant Silene acaulis as foundation species: a bug’s-eye view to facilitation and microclimate. PLoS One. 2011; 7(5): e37223. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorgan MT, Wilson WG, Knight TM: Plant population dynamics, pollinator foraging, and the selection of self-fertilization. Am Nat. 2005; 166(2): 169–183. PubMed Abstract \| Publisher Full Text\n\nMorris WF, Doak DF: Life history of the long-lived gynodioecious cushion plant Silene acaulis (Caryophyllaceae), inferred from size-based population projection matrices. Am J Bot. 1998; 85(6): 784–793. PubMed Abstract \| Publisher Full Text\n\nMuñoz AA, Arroyo MTK: Pollen limitation and spatial variation of reproductive success in the insect-pollinated shrub Chuquiraga oppositifolia (Asteraceae) in the Chilean Andes. Arct Antarc Alp Res 2006; 38(4): 608–613. Publisher Full Text\n\nPojar J, Klinka K, Meidinger DV: Biogeoclimatic ecosystem classification in British Columbia. For Ecol Manage. 1987; 22(1–2): 119–154. Publisher Full Text\n\nPotts SG, Biesmeijer JC, Kremen C, et al.: Global pollinator declines: trends, impacts and drivers. Trends Ecol Evol. 2010; 25(6): 345–353. PubMed Abstract \| Publisher Full Text\n\nPrimack RB: Variability in New Zealand montane and alpine pollinator assemblages. N Z J Ecol. 1978; 1: 66–73. Reference Source\n\nPrimack RB: Insect pollination in the New Zealand mountain flora. N Z J Ecol. 1983; 21(3): 317–333. Publisher Full Text\n\nReid AM, Hooper R, Molenda O, et al.: The reproductive effects of reduced pollen deposition via exclosures and hand pollination on the cushion plant Silene acaulis. 2014. Data Source\n\nReid AM, Lortie CJ: Cushion plants are foundation species with positive effects extending to higher trophic levels. Ecosphere. 2012; 3(11). Publisher Full Text\n\nRichards AJ: Plant Breeding Systems: Second edition. Chapman & Hall, London. 1997. Reference Source\n\nSAS. Jmp SAS Institute Inc. 2012. Reference Source\n\nSchöb C, Michalet R, Cavieres L, et al.: A global analysis of bidirectional interactions in alpine plant communities show facilitators experiencing strong reciprocal fitness costs. New Phytol. 2014; 202(1): 95–105. PubMed Abstract \| Publisher Full Text\n\nShykoff JA: Maintenance of gynodioecy in Silene acaulis (Caryophyllaceae): stage-specific fecundity and viability selection. Am J Bot. 1988; 75(6): 844–850. Publisher Full Text\n\nShykoff JA: Sex polymorphism in Silene acaulis (Caryophyllaceae) and the possible role of sexual selection in maintaining females. Am J Bot. 1992; 79(2): 138–143. Publisher Full Text\n\nTorres-Díaz C, Gómes-González S, Stotz GC, et al.: Extremely long-lived stigmas allow extended cross-pollination opportunities in a high Andean plant. PLoS One. 2011; 6(5): e19497. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTotland Ø: Pollination in alpine Norway: Flowering phenology, insect visitors, and visitation rates in two plant communities. Can J Bot. 1993; 71(8): 1072–1079. Publisher Full Text\n\nTotland Ø, Sottocornola M: Pollen limitation of reproductive success in two sympatric alpine willows (Salicaceae) with contrasting pollination strategies. Am J Bot. 2001; 88(6): 1011–1015. PubMed Abstract \| Publisher Full Text\n\nUtelli AB, Roy BA: Pollinator abundance and behavior on Aconitum lycoctonum (Ranunculaceae): an analysis of the quantity and quality components of pollination. Oikos. 2000; 89(3): 461–470. Publisher Full Text" }	[ { "id": "5191", "date": "02 Jul 2014", "name": "Fabio Bulleri", "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 an interesting article reporting on a study investigating how reduced pollen supply influences reproduction of self-compatible versus self-incompatible plants of an alpine cushion species and its interaction with associated species. The main strength of the study is that of providing hints on the potential role that factors other than environmental conditions can play in modulating species interactions. In a relatively recent work by Maestre et al. (2009), stress has been divided into two major types, resource and non-resource. I am not sure whether pollen availability can be seen as a resource, but I would not have major issues with viewing pollinator availability as a quantifiable, limiting resource. Maybe, these concepts could be touched upon somewhere in the Introduction.The main weakness of the study is the lack of quantification of pollen delivery in different treatments: I think the authors should try to give the reader, at the very least, a gross idea of the order of magnitude of differences between natural and artificial pollen supply. The second point is that there is not a formal gradient of pollen supply, as this implies more than two levels. This does not impinge on the robustness of the results, but limits the insight the study offers into the Stress Gradient Hypothesis.Finally, both in the Abstract and M&M I did not get clearly who the benefactors and beneficiary were. Actually, only from the Results did I understand that you were analyzing the effects of species hosted by the cushion on the cushion-forming species. I sense that referring to the cushion and associated species might be a better solution.SPECIFIC COMMENTSIntroductionIt is not clear what grounds the alternative model for pollinator abundance (increasing with elevation) in alpine environments builds upon. In the Introduction, you refer to the stress source sometimes as ‘pollination supply’ and sometimes as ‘reduced pollen’ or ‘less pollen’. It would be useful to specify if the limiting resource is pollen supply (which implies reduced production by plants) or pollinator abundance (I guess the second, from the model you propose in the Introduction). Is there any evidence of competition between hermaphrodite and female plants when pollen is not limiting? Material & MethodsYou need to be consistent in the way you refer to hermaphrodite plants. Some plants could receive pollen from insects while others were covered with mesh and hand-pollinated. It is not clear how insect versus hand pollination generated a gradient of pollen supply. Do you assume that delivery of pollen by insect was much higher? If so, which order of magnitude? “Instead of testing the effect of beneficiaries on measures of reproductive success over a range of environmental gradients (Bertness & Callaway, 1994) we tested whether variation in the current and experimentally reduced pollen supply can be viewed as potential stressor.” Again, it is not clear here who is expected to be facilitated and by whom (see my general comments). Likely, it would be useful to be more specific in your predictions at the end of the Introduction. Also, switch beneficiary and % fruit set in the last sentence of the Abstract (the way it is, it suggests that you are mainly looking at effect of fruit set on beneficiary abundance).ResultsEstimates of effects could be reported in Tables, so as to give an idea of their direction. The relationship with beneficiary abundance explains little variation in % fruit set (Fig. 3), suggesting that other factors are of overwhelming importance. This point should be made clear in the Discussion.DiscussionIt would be interesting if the authors could offer a possible explanation for a positive relationship between beneficiary abundance and % fruit set under reduced pollen supply. What are the potential mechanisms?", "responses": [ { "c_id": "889", "date": "02 Jul 2014", "name": "Christopher Lortie", "role": "Reader Comment", "response": "Thank you for the suggestions. We will address all of them and provide a revised version." }, { "c_id": "895", "date": "08 Jul 2014", "name": "Christopher Lortie", "role": "Author Response", "response": "Our response to Dr Bulleri’s comments:I am not sure whether pollen availability can be seen as a resource, but I would not have major issues with viewing pollinator availability as a quantifiable, limiting resource. Maybe, these concepts could be touched upon somewhere in the Introduction.This is a good point. This is now added in the Abstract and made clearer in the introduction. I think the authors should try to give the reader, at the very least, a gross idea of the order of magnitude of differences between natural and artificial pollen supply.Unfortunately this is very difficult as there is no estimation of current pollen deposition rates. This limitation has been more clearly stated now in the following sentence, “It is assumed that plants that are covered and hand pollinated three times receive less pollen than the plant open to natural pollination during the 33 snow-free days”. The second point is that there is not a formal gradient of pollen supply, as this implies more than two levels. This does not impinge on the robustness of the results, but limits the insight the study offers into the Stress Gradient Hypothesis.Agreed. This is clearly stated in the following sentence, “Instead of testing the effect of beneficiaries on measures of reproductive success over a range of environmental gradients (Bertness and Callaway 1994) we test this effect under two pollen regimes: current and reduced”. It is not clear what grounds the alternative model for pollinator abundance (increasing with elevation) in alpine environments builds upon.Good question. Unfortunately, there is really no model for pollinator abundance in the alpine/on elevation gradients. We would love to develop one though but likely not in the context of this outcrossing experiment as it is focused on botany and ecology. We hope that the proposal that pollen limitation is a potential stressor can lead to the expansion of this set of theory. Nice opportunity for the future. In the Introduction, you refer to the stress source sometimes as ‘pollination supply’ and sometimes as ‘reduced pollen’ or ‘less pollen’. It would be useful to specify if the limiting resource is pollen supply (which implies reduced production by plants) or pollinator abundance (I guess the second, from the model you propose in the Introduction).Yes, great. Thank you. Is there any evidence of competition between hermaphrodite and female plants when pollen is not limiting?Yes. See Figure 3 and the following sentence in the results, “Under the current pollination rates, percent fruit-set and beneficiary abundance are negatively related (slope = -0.48, R2 = 0.04, Figure 3).”Material & MethodsYou need to be consistent in the way you refer to hermaphrodite plants.Revised. Some plants could receive pollen from insects while others were covered with mesh and hand-pollinated. It is not clear how insect versus hand pollination generated a gradient of pollen supply. Do you assume that delivery of pollen by insect was much higher? If so, which order of magnitude?Yes, the assumption is that the plants that had pollen delivered by insects received more pollen because they could receive pollen every day compared to the three hand-pollination events. The order of magnitude of this difference would be purely a guess. This is a very important assumption and limitation of this study. Again, it is not clear here who is expected to be facilitated and by whom (see my general comments). Likely, it would be useful to be more specific in your predictions at the end of the Introduction.Revised, thank you. “Specifically, we predict facilitation between cushions and beneficiaries to increase when pollen supply is reduced.” Also, switch beneficiary and % fruit set in the last sentence of the Abstract (the way it is, it suggests that you are mainly looking at effect of fruit set on beneficiary abundance).Revised and clearer now.ResultsEstimates of effects could be reported in Tables, so as to give an idea of their direction.Good idea Fabio. We chose to do stats on the primary dataset and not calculate derived data such as effect size measures (Rii) for two reasons. We were not confident that Rii, now common in facilitation studies, was appropriate to the treatments here and secondly that we elected to visualize percent change in Figure 1. We feel that figure one does a good job of showing the strength of evidence, i.e. effects, and that a table was thus not needed. Let us know if you think we should also include the data from this figure in a table but we think the current figure gives a good overview of the evidence/biological significance. The relationship with beneficiary abundance explains little variation in % fruit set (Fig. 3), suggesting that other factors are of overwhelming importance. This point should be made clear in the Discussion.The following sentence has been added to the discussion, “That beneficiary abundance explains little variation in percent fruit-set (low correlation coefficients), suggests that other variables are of greater importance in determining percent fruit-set.”DiscussionIt would be interesting if the authors could offer a possible explanation for a positive relationship between beneficiary abundance and % fruit set under reduced pollen supply. What are the potential mechanisms?Good point. I think it may be too premature to assert possible reasons for this response. We do highlight the need for further research on this topic in the sentence, “Clearly, additional research is needed to identify the causal mechanisms between plant-plant interactions and plant-pollinator interactions with dominant plant species that host other species.”" } ] }, { "id": "5192", "date": "02 Jul 2014", "name": "He Qiang", "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 conducted experiment and well written manuscript. It is the first experimental test of the stress-gradient hypothesis under pollen limitation stress, showing how we can further our understanding of plant-plant interactions with a new perspective. In my view, this is a neat paper, though I have some concerns that I hope the author can address in a revision.The main limitation of this paper is that plant-plant interactions are inferred by association, not experimental removal of neighbours. Although a number of studies from alpine habitats suggest that facilitation can be indicated by positive association relationships, this limitation I feel should be clarified in the paper.It is also a bit confusing to me why the authors did hand pollination for reduced-pollination treatment plants - was this because reduced-pollination treatment plants did not receive pollination at all without hand pollination? If so, the question would be how the authors controlled the magnitude of hand pollination as compared to natural pollination.A third issue is that the pollen reduction treatments lacked procedural control. For example, one cannot tell the effect of mesh on the data. Also surely, the authors need to provide some details about the cloth mesh they used, such as mesh size etc.It is unclear what can be the potential mechanisms for facilitation under pollen limitation. Facilitation in physically stressful habitats is produced by neighbours’ relief of abiotic stress, and in biologically stressful habitats by associational defence. It is not a must for the authors to experimentally demonstrate this in the paper but I think they should at least discuss some of the possibilities. The authors may also want to include descriptions to encourage future research that experimentally tests for these possibilities.I have several more specific comments:\"Before bud-burst ...\" Clarify the start and end dates of your experimental treatments. What is the total duration? \"The first 20 female and male plants found were ...\" Does the male here mean \"hermaphrodite\"? Use terms consistently. For treatments, how many plants in current pollination treatments are female and male, respectively? \"Reproductive output measures were collected from the 40 ...\" 40 hand pollinated, 231 naturally pollinated, this is 271 in total. So this number is inconsistent with the numbers described above: 273 plants were measured. There are three plants that were excluded? Clarify when the beneficiary abundance was quantified and how this was done. It is the total number of plants living within the cushion. Shouldn't it be density per unit area? How many beneficiary species occurred within your 273 study cushions? Would it be possible that beneficiary diversity rather than abundance is more likely to affect cushion plant reproductive success? Actually, the beneficiary increased cushion plant fruit set under reduced pollen limitation. This means the beneficiary was actually a benefactor. This is a bit confusing; might it use neighbours instead? \"we used a generalized linear model (GLM) with Poisson distribution ...\" Poisson distribution is often used for positive integer response variables. How about over-dispersion? Over-dispersion can substantially affect significance tests. \"Under the current pollination rates, percent fruit-set and beneficiary abundance are negatively related ...\" Did pollination reduction affect the relationship between other reproductive variables and beneficiary abundance?Overall, this is a great paper. The publication of this paper can be a stimulator of a new direction of positive species interactions with environmental stress.", "responses": [ { "c_id": "888", "date": "02 Jul 2014", "name": "Christopher Lortie", "role": "Reader Comment", "response": "Fantastic feedback. Thank you. We will address all these comments and revise appropriately." }, { "c_id": "894", "date": "07 Jul 2014", "name": "Christopher Lortie", "role": "Author Response", "response": "Our response to Dr He’s comments:The main limitation of this paper is that plant-plant interactions are inferred by association, not experimental removal of neighbors. Although a number of studies from alpine habitats suggest that facilitation can be indicated by positive association relationships, this limitation I feel should be clarified in the paper.The following sentence was added to the ‘Data analysis’ section, “As there was no experimental removal of beneficiaries, interactions are inferred by association”. ...how the authors controlled the magnitude of hand pollination as compared to natural pollination.The following sentences were added to the ‘Treatment’ section, “To experimentally reduce plant pollination, plants were covered with mesh and hand pollinated three times over the duration of the snow-free season. It is assumed that plants that are covered and hand pollinated three times receive less pollen than the plant open to natural pollination during the 33 snow-free days.” A third issue is that the pollen reduction treatments lacked procedural control. ...Also surely, the authors need to provide some details about the cloth mesh they used, such as mesh size etc.It is not really possible to put mesh on without influencing pollinators so this was not conducted. We used fine cloth mesh purchased from a fabric store. It was flexible and durable. This was added to the methods section. See reference Donnelly et al. 1998. It is unclear what can be the potential mechanisms for facilitation under pollen limitation. ...It's not a must for the authors to experimentally demonstrate this in the paper. But I think at least they should discuss some of the possibilities. The authors may also want to include descriptions to encourage future research that experimentally tests for these possibilities.Good point. I think it may be too premature to assert possible reasons for this response. We do highlight the need for further research on this topic in the sentence, “Clearly, additional research is needed to identify the causal mechanisms between plant-plant interactions and plant-pollinator interactions with dominant plant species that host other species.”Responses to specific comments:Added from June 26th to August 29th, 2010.Yes, great point. Revised.Added; 124 were female and 107 were hermaphrodites.A total of 271 S. acaulis plants were measured.Yes, but we do not have % area covered unfortunately. We took measures of beneficiary abundance and richness.We checked beneficiary diversity and found that it was not strongly related, but that beneficiary abundance was a more informative measure. There were 16 other plant species growing on the cushions.I think introducing a new term would cause confusion not add clarity. They are the same plants they just appear to switch roles (competitor to facilitator) depending on pollen supply.The over-dispersion tests in the GLM did change some of the p-values of the factors but not near the alpha of p < 0.05.This pattern is consistent in the measure of seeds/fruit but not the other measures of reproductive output." } ] }, { "id": "5194", "date": "04 Jul 2014", "name": "Ernesto Badano", "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 dealing with the effects that facilitated plants may have on the performance of the reproductive success of the nurse species. As far as I am aware, this is the first study reporting such effects. The article is well written and is easy to read. The arguments provided in the introduction section support the hypotheses raised by the authors, and the methods used to test these hypotheses are sound. I believe that this paper makes an important contribution to the conceptual framework of positive interactions among plants along environmental gradients, mainly because it helps one to understand the outcome of these interactions in future climate change scenarios. I just have a few comments and queries about the methods used by the authors. The other sections of the article are clear and would not require major improvement. Materials and Methods: Specific commentsIn the subsection “Study species”, several sentences begin with the abbreviated scientific name of species (S. acaulis). Nevertheless, it is not appropriate to use the abbreviated name of the species for beginning a sentence. In this case, the full name of the species should be provided (Silene acaulis) while its abbreviated form (S. acaulis) must be used when the species is mentioned in the middle of a sentence. In the second paragraph of the subsection “Study species” the authors indicate that, besides S. acaulis, several other plant species are relatively abundant. My question on this issue is: Are all these other species beneficiary of S. acaulis or they are just sharing the habitat with this cushion plant? This should be clarified to understand the aim of testing the Stress Gradient Hypothesis. In the first paragraph of the subsection “Treatments” the authors indicate that they choose 20 female and male plants for the experiment. This caught my attention because in the previous sections of the article the authors stated that they used a gynodioecious population of S. acaulis and, therefore, male plants should not exist. I think that there is a mistake in this sentence, where instead of “20 female and male plants” the authors should say “20 female plants and plants with hermaphroditic flowers”.", "responses": [ { "c_id": "893", "date": "07 Jul 2014", "name": "Christopher Lortie", "role": "Author Response", "response": "Our response to Dr Badano’s comments:Completed. Thank you. These species were present both on and beside S. acaulis cushions. This is now clarified. Yes, you are correct. This is now fixed." } ] } ]	1	https://f1000research.com/articles/3-130
https://f1000research.com/articles/3-186/v1	06 Aug 14	{ "type": "Case Report", "title": "Case Report: A case of antiphospholipid-antibody-associated reactive angioendotheliomatosis and panniculitis", "authors": [ "Emily D. Privette", "Campbell L. Stewart", "Steffan W. Schulz", "Victoria P. Werth", "Emily D. Privette", "Campbell L. Stewart", "Steffan W. Schulz" ], "abstract": "Introduction: Reactive angioendotheliomatosis (RAE) is a benign vascular disorder with a varied clinical presentation that has been associated with a wide range of systemic illnesses. It is characterized histologically by intravascular and extravascular hyperplasia of endothelial cells and pericytes, and this entity only affects the skin. We present the rare case of a woman with both RAE and panniculitis associated with anticardiolipin antibodies in the setting of antiphospholipid syndrome (APLS). Case description: A female in her forties presented for a second opinion of lumps in her skin. The first lump was noted on her right thigh four years prior to presentation, and biopsy demonstrated longstanding septal and lobular panniculitis. Studies that followed this finding revealed a diagnosis of APLS, and the patient had been placed on 400 mg per day of hydroxychloroquine and a daily low-dose of aspirin. She presented with a new subcutaneous plaque on the left lower back, and biopsy of this lesion was suggestive of RAE. She had no history of significant thrombotic events and had never been treated with an anticoagulant. Work-up was negative for other connective tissue diseases that are known for producing subcutaneous nodular lesions. Discussion: RAE is very rare, with less than 50 cases reported in total; five of these in association with APLS. In the context of APLS, RAE is likely a result of the procoagulant state induced by the presence of antiphospholipid antibodies. This case of both RAE and panniculitis associated with anticardiolipin antibodies in the setting of APLS describes rare manifestations of APLS as isolated incidents in a relatively healthy patient.", "keywords": [ "The patient was a 49 year-old Caucasian female legal assistant", "with a history of mild hypertension", "taking hydrochlorthiazide and metoprolol. Chronic leg pain", "panniculitis without known primary explanation", "and antiphospholipid antibody syndrome (APLS) presenting for a second opinion of lumps in her skin." ], "content": "Case\n\nThe patient was a 49 year-old Caucasian female legal assistant, with a history of mild hypertension, taking hydrochlorthiazide and metoprolol. Chronic leg pain, panniculitis without known primary explanation, and antiphospholipid antibody syndrome (APLS) presenting for a second opinion of lumps in her skin.\n\nThe first lump had been noted on her right thigh four years prior to presentation, when the patient noticed darkening of the overlying skin, giving it a bruise-like appearance, with a central nodule. The area itself was non-tender but was associated with lateral leg pain that, according to the patient, failed to respond to physical therapy and multiple cortisone injections. Biopsy of the thigh nodule demonstrated features of a longstanding septal and lobular panniculitis. The nodule was surgically removed and did not recur. After excision of the nodule, the patient’s leg pain decreased, though she felt some residual throbbing, especially at night and with sitting. Approximately six months later she developed two other subcutaneous lumps, on the back and left-lower abdomen. Biopsy of the abdominal lesion was consistent with a lipoma.\n\nReview of systems was unremarkable, as there were no joint pains or swelling of the joints, and no fevers, chills, night sweats, or significant weight change.\n\nPast medical history included hypertension and hypercholesterolemia. The patient had no history of significant thrombotic events and had never been treated with an anticoagulant.\n\nPrior lab investigations, approximately two years prior to presentation in our clinic, revealed an Antinuclear Antibody Test (ANA) that stained positive at dilutions of 1:160 and 1:320, positive anticardiolipin IgM antibody at 93 MPL, and positive Beta-2-glycoprotein 1 IgM antibody at a level greater than 150 MPL. A hematologic panel and coagulation studies were within normal limits. Rheumatologic work-up was also negative for rheumatoid factor, SSA/SSB, and smooth muscle or RNP antibodies. Upon diagnosis of APLS the patient had been placed on 400 mg per day of hydroxychloroquine and 81 mg per day of aspirin.\n\nOn exam there was a firm subcutaneous plaque on the left lower back, approximately 3×3 cm, with no overlying skin changes (Figure 1), a firm, mobile nodule, approximately 2×2 cm on the left lower quadrant of the abdomen, and a livedo on the extensor surfaces of the upper arms and legs.\n\nThis photo shows the firm subcutaneous plaque on the patient’s left lower back, approximately 3×3 cm, with no overlying skin changes. On microscopic exam this lesion was found to be RAE.\n\nA biopsy of the patient’s back lesion was reviewed by a dermatopathologist and described as a proliferation of spindled endothelial cells and vessels with concentric wall thickening in the subcutaneous tissue (Figure 2). The spindled cells were highlighted with a CD31 IHC staining (Figure 3), and a Factor XIIIa staining also highlighted many cells within the spindle cell proliferation, confirming their endothelial origin. Additionally, within the dermis there was a perivascular and periadnexal lymphocytic infiltrate. Neither vascular occlusion nor microvascular thrombosis was observed.\n\nOverall, the spindle cell proliferation within the subcutaneous tissue was thought to be suggestive of a reactive angioendotheliomatosis (RAE). In response to this result, the patient’s low dose aspirin and hydroxychloroquine regimen was continued, and she was advised to discontinue oral contraceptive pills (OCPs) given the concern for clotting risk. The patient discontinued OCPs and opted not to have her remaining lesions surgically removed. At six months follow-up, the current lesions have stabilized and no new lesions have developed.\n\nAn H&E staining of the biopsy section of the patient’s back lesion demonstrating proliferation of spindled endothelial cells and vessels with concentric wall thickening in the subcutaneous tissue. 10× magnification.\n\nA CD31 immunohistochemical staining of the patient’s back lesion highlighting the spindled cells. 10× magnification.\n\n\nDiscussion\n\nRAE is a benign vascular disorder with a varied clinical presentation that has been associated with a wide range of systemic illnesses. It is characterized histologically by intravascular and extravascular hyperplasia of endothelial cells and pericytes, and this entity only affects the skin1. The intravascular cells are without atypia and have been demonstrated to display reactivity for antibodies to Factor VIII-related antigen, blood group isoantigens A, B, and H, vimentin, and Ulex europaeus I lectin, and negativity for leukocyte common antigen (LCA)2. RAE is very rare, with less than 50 cases reported1. In contrast, malignant angioendotheliomatosis (MAE), a fatal intravascular lymphoma of B-cell origin, displays reactivity for antibodies to LCA, B-Cell antigens, and vimentin in tumor cells2.\n\nOther systemic illnesses that have been associated with RAE include subacute bacterial endocarditis, tuberculosis, lymphoproliferative disorders, liver failure, renal failure, rheumatoid arthritis, cryoglobulinemia, and peripheral vascular atherosclerosis3.\n\nClinically, the presentation of RAE can vary widely, ranging from erythematous or purpuric papules, macules and plaques to ecchymoses, which may exhibit necrosis or ulceration4. The lesions have been observed to mimic the following diagnoses: Kaposi sarcoma, morphea, pyoderma gangrenosum, calciphylaxis, angiosarcoma, lichen ruber verrucosus, sarcoid, pyogenic granuloma, eruptive disseminated lobular capillary haemangioma, and lupus panniculitus4. The clinical appearance of our patient’s back lesion, an indurated, subcutaneous plaque with no change to the overlying skin, led to a differential of lupus panniculitis versus morphea prior to histological diagnosis.\n\nThe pathogenesis of RAE remains unclear, although the fact that it is associated with various disparate disease entities suggests multiple pathways leading to a common reactive pattern3. Additionally, RAE occurs in the context of conditions that may cause occlusion of the vascular lumina. This has led to the suggestion that vascular occlusion by thrombi may cause localized hypoxia and acidosis, followed by hyperplasia of endothelial cells and occasionally pericytes, leading to the observed histopathology5.\n\nIn this case, it is likely that the findings of RAE and panniculitis are both associated with the patient’s APLS. Specifically, the pathogenesis of these cutaneous findings has been associated with the presence of anticardiolipin antibodies, which were present in our patient. Localized RAE in the setting of APLS, such as our patient’s back lesion, has been previously described in the literature3, with the first known case reported in 20006 and less than 5 cases reported in total3,5,6. In the context of antiphospholipid syndrome, RAE is likely a result of the procoagulant state induced by the presence of antiphospholipid antibodies. A recent report postulates that the anticardiolipin antibody may be the underlying cause for the thrombophilia leading to the finding of RAE, given that a combination of microthrombosis and reactive endothelial-cell proliferation has been reported in association with anticardiolipin antibodies in the literature5. The patient’s thigh nodule biopsy, which demonstrated panniculitis is also consistent with her diagnosis of APLS. Anticardiolipin antibodies have been associated with septal and lobular panniculitis resembling the features noted in the patient’s biopsy7. Additionally, previous cases have also reported the association of patients with APLS, thigh nodules with leg pain, and elevated anticardiolipin antibody levels7,8. It is unlikely that these cutaneous findings are related to another primary disorder, given the patient’s negative work-up and lack of symptoms indicative for systemic lupus erythematous (SLE), sarcoidosis, fascitis, or scleroderma/morphea, polyarteritis nodosa, or other connective tissue diseases known for producing subcutaneous nodular lesions.\n\nIn conclusion, we report the rare case of a woman with both RAE and panniculitis associated with anticardiolipin antibodies in the setting of APLS. This case adds to the literature by describing these rare manifestations of APLS as isolated incidents in a relatively healthy patient.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nVW and SS identified this interesting case that warranted reporting. EP prepared the first draft of the manuscript. CS supplied and photographed figures. 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 project is supported by the Department of Veterans Affairs Veterans Health Administration, Office of Research and Development, Biomedical Laboratory 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\nRongioletti F, Rebora A: Cutaneous reactive angiomatoses: patterns and classification of reactive vascular proliferation. J Am Acad Dermatol. 2003; 49(5): 887–896. PubMed Abstract \| Publisher Full Text\n\nWick MR, Rocamora A: Reactive and malignant “angioendotheliomatosis”: a discriminant clinicopathological study. J Cutan Pathol. 1988; 15(5): 260–271. PubMed Abstract \| Publisher Full Text\n\nThai KE, Barrett W, Kossard S: Reactive angioendotheliomatosis in the setting of antiphospholipid syndrome. Australas J Dermatol. 2003; 44(2): 151–155. PubMed Abstract \| Publisher Full Text\n\nMcMenamin ME, Fletcher CD: Reactive angioendotheliomatosis: a study of 15 cases demonstrating a wide clinicopathologic spectrum. Am J Surg Pathol. 2002; 26(6): 685–697. PubMed Abstract\n\nKirke S, Angus B, Kesteven PJ, et al.: Localized reactive angioendotheliomatosis. Clin Exp Dermatol. 2007; 32(1): 45–47. PubMed Abstract \| Publisher Full Text\n\nCreamer D, Black MM, Calonje E: Reactive angioendotheliomatosis in association with the antiphospholipid syndrome. J Am Acad Dermatol. 2000; 42(5 Pt 2): 903–906. PubMed Abstract \| Publisher Full Text\n\nHunt RD, Robinson M, Patel R, et al.: Antiphospholipid-antibody-associated panniculitis. Dermatol Online J. 2012; 18(12): 18. PubMed Abstract\n\nRenfro L, Franks AG Jr, Grodberg M, et al.: Painful nodules in a young female. Antiphospholipid syndrome. Arch Dermatol. 1992; 128(6): 847–8. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5938", "date": "27 Aug 2014", "name": "Anne E. Laumann", "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\nReviewing this paper has taught me about reactive angioendotheliomatosis, a condition about which I knew little before. As such this is a relevant report for indexation. However, it can be made more concise and some of the grammar can be corrected.Abstract is excellent, except aspirin was being used as an anticoagulant.Some suggestions:Para 1: No reason for “The patient was ---” and “ ---with a history of mild hypertension, taking hydrochlorothiazide and metoprolol”. Instead: “A 49 year old female legal assistant with chronic leg pain, panniculitis of unknown cause, and antiphospholipid antibody syndrome (APLS) presented for a further opinion related to lumps in her skin”Para 3: Review of systems revealed no pain or swelling of her joints, no fever, chills, night sweats nor weight change.Para 4: She had known hypercholesterolemia and hypertension, for which she was taking metroprolol and hydrochlorothiazide. She had had no previous significant thrombotic events, but to say she had never had an anticoagulant is not really correct as she was on daily aspirin and in this situation the hydroxychloroquine was being used to help decrease blood viscosity also. May be “anticoagulant” is a question of definition.Para 5: MPL needs defining and the units for Beta-2-glycoprotein need correcting.Sentence 3: suggest: Rheumatoid factor, SSA/SSB and smooth muscle and RNP antibodies were negative.Para 6: delete “a” prior to “livedo”Para 7: \"a dermatopathologist\" – presumably one of the authors. IHC needs definingDiscussionPara 1: Sentence 3: \"without atypia and display\" - delete “have been demonstrated to”Para 3: Sentence 2: “Panniculitus” needs correctingPara 5: Sentence 1: delete “patient’s” in –“ In this case, it is likely that the findings of RAE and panniculitis are both associated with the APLS.”Sentence 3:\n\ndelete “in the literature”. Change: “report” rather than “reported” in “with the first known case reported in 2000”Sentence 5: delete “in the literature”Sentence 7: remove “patient’s”Sentence 8: Only “Additionally” or “also” is needed, not bothSentence 9: remove “patient’s”. Why do you need “(SLE)” as this is not mentioned again? “Fascitis” should be “fasciitis”.", "responses": [] }, { "id": "5732", "date": "01 Sep 2014", "name": "Beatrix Volc-Platzer", "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 view of the rarity of reactive angioendotheliomatosis this is an important case report to alert clinicians to this disorder and should eventually, be indexed. However, I think that the authors should go into more detail as listed below, and I therefore approve with reservations at the moment. Abstract:Introduction: It is not only the clinical presentation that is quite varied, but also the histopathological pattern. Intravascular endothelial proliferation is regarded as \"key feature\", but other patterns have been described as well. Deposition of fibrin and thrombi are variably found within the endothelial proliferations. Perivascularly, extravasation of erythrocytes and lymphocytes are found... therefore, I suggest adapting the second sentence of the introduction accordingly. Case description:Regarding the case description I have a few suggestions:A female patient ....... (APLS) due to ....The dose of aspirin should be given as well.It should be stated also in the abstract whether - in addition to connective tissue diseases - other potential stimuli of or disease associations with RAE, respectively, have been excluded, e.g. endocarditis, liver or renal disease. Case Report:Case:Regarding the case I have a few questions:The medication for hypertension is given, but not whether hypercholesterolemia has been treated and, if so, by which medication.What was the pattern of ANA staining? In case of positive ANA and in particular with a homogenous pattern, anti-dsDNA-antibodies were negative?What are the reference values for antiphospholipid IgM and ß2-glycoprotein 1?Have cryoglobulins been tested for?I suggest to mention in the text the reason for using hydroxychloroquine (antithrombotic and antihyperlipaemic) in addition to aspirin.Figure 1: Is the central pale area a scar after biopsy?Histopathology: as far as I see in Figure 2 the main pathology is at the dermal/subcutaneous interface. RAE, however, is found in the superficial dermis or throughout the entire dermis. Histopathology reminds me rather of panniculitis. Is it possible to find another figure showing more of the dermis?Immunohistochemical studies have been performed with anti-CD34, vWF, smooth muscle actin for pericytes, and HHV-8.Discussion:Discussion includes all important features of RAE. However, discussion on histopathological patterns might be extended.1st paragraph: ... on various forms of RAE, i.e. diffuse dermal angiomatosis - which might be the main histopathologic differential diagnosis, glomeruloid hemangioma-like, angiomatosis with deposits of cryoglobulin, etc.", "responses": [] } ]	1	https://f1000research.com/articles/3-186
https://f1000research.com/articles/3-42/v1	13 Feb 14	{ "type": "Research Article", "title": "Nominal Group Technique consultation of a Pulmonary Rehabilitation Programme", "authors": [ "Hayley A Hutchings", "Frances L Rapport", "Sarah Wright", "Marcus A Doel", "Clare Clement", "Keir E Lewis", "Frances L Rapport", "Sarah Wright", "Marcus A Doel", "Clare Clement", "Keir E Lewis" ], "abstract": "Objective: The purpose of the study was to determine what patients, professionals and significant others regarded as the most important positive- and challenging aspects of Pulmonary Rehabilitation Programmes for patients with Chronic Obstructive Pulmonary Disease (COPD) and to gain insight into how such programmes could be developed and improved.Method: A modified Nominal Group Technique method was used in three consultation workshops (one with COPD patients who had recently undertaken a Pulmonary Rehabilitation Programme; one with ‘significant others’ of the same patients; one with secondary care professionals who deliver the Pulmonary Rehabilitation Programme).Results: Each of three workshops resulted in the production of approximately ten positive- and ten challenging aspects related to Pulmonary Rehabilitation Programmes. These were further developed by a process of thematisation into seven broad themes. The most important was ‘the patient’, followed by ‘physical health’; jointly ranked as third were: ‘mental health’ and ‘knowledge and education’. ‘The programme’ and ‘professional characteristics’ were jointly ranked as fifth, with ‘the future’ being ranked as the least important theme.Conclusions: The modified Nominal Group Technique method allowed the development of a ranked thematic list that illustrated the important positive- and challenging aspects of Pulmonary Rehabilitation Programmes for patients with COPD. These themes should be core to planning future Pulmonary Rehabilitation Programmes, particularly if patients and carer views are to be considered.", "keywords": [ "Chronic Obstructive Pulmonary Disease", "patients", "professionals", "Pulmonary Rehabilitation Programme", "Nominal Group Technique" ], "content": "Introduction\n\nChronic Obstructive Pulmonary Disease (COPD) is a progressively disabling condition characterised by impaired respiratory function associated with physical limitations and psychological co-morbidity1. COPD results in a reduced capacity for functional activities and performance of daily activities with a corresponding impairment in Health Related Quality of Life2. Current figures show 900,000 people have been diagnosed with and are receiving treatment for COPD within the United Kingdom3. However, due to under reporting or under diagnosis, the actual number of those suffering with COPD is estimated to be as high as 3 million4. Stopping smoking is crucial and is the only intervention that influences the natural history of lung deterioration, with current pharmacological treatment being aimed at reducing symptoms and exacerbations5.\n\nPulmonary Rehabilitation Programmes are multi-disciplinary interventions individually tailored to optimise each patient’s physical and social performance. Rigorous evidence from randomised controlled trials demonstrates that Pulmonary Rehabilitation Programmes for COPD can improve dyspnoea, exercise tolerance, Health Related Quality of Life, and reduce the number of days spent in hospital and the utilisation of healthcare resources6–8. Pulmonary Rehabilitation Programmes have been shown to be cost-effective and are now recommended for all patients who remain breathless despite optimal bronchodilators, irrespective of severity and age6–9. Pulmonary Rehabilitation Programmes are also being effectively applied to non-COPD causes of pulmonary impairment10.\n\nThere are now specific guidelines and recommendations in the United Kingdom regarding Pulmonary Rehabilitation Programmes, including how to select patients, the timing and number of sessions, intensity and type of exercise, the key educational, psychological and behavioural components, oxygen supplementation and outcome assessment7,8. Research exploring the benefits following Pulmonary Rehabilitation Programmes has predominantly been quantitative in nature. There have been some qualitative studies with COPD patients, but these have focused largely on specific aspects of patient experience11,12 and barriers to participation in Pulmonary Rehabilitation Programmes or other self-management programmes13,14. There has been some exploration of the effectiveness of self-management programmes from the patient perspective15–17. However, none of these studies have combined patient, carer, and professional perspectives, particularly in an in-depth analysis regarding the long-term impact of Pulmonary Rehabilitation Programmes in relation to personal needs and issues such as perceived patient benefits, and expectations and challenges of Pulmonary Rehabilitation Programmes. It has been recognised that a better understanding of how Pulmonary Rehabilitation Programmes improve Health Related Quality of Life could affect the future design of programmes, enhance measurement tools for Health Related Quality of Life and more appropriately support the specific needs of patients15,17,18.\n\nConsensus methods are techniques used gain opinions and views from appropriate experts regarding the current position in a particular field. They provide a mechanism for assimilating and synthesising information, particularly where published information may be inadequate or non-existent19. The purpose of consensus methods is to reach an agreement on a particular issue. Consensus methods can also mitigate some of the problems sometimes associated with group decision-making processes. In particular, where dominant views may lead the process and crowd out other perspectives19.\n\nNominal Group Technique is one of the commonly used consensus methods within healthcare and medical settings. The technique was first developed as an organisational planning technique by Delbecq et al. in the 1970s20. The Nominal Group Technique normally involves four main phases: a nominal phase, during which each individual silently considers the issues under deliberation; an item-generation phase, during which each individual discloses the results of their deliberation to the group; a discussion and clarification phase, during which the group assures itself that it has understood the items that have been advanced; and a voting phase, during which the items are evaluated and the issue is decided (e.g. a ranking exercise). Nominal Group Technique promotes individual contributions allowing each individual the opportunity to voice their opinions. Factors that would normally inhibit participation are therefore avoided and even the more reticent group members are encouraged to participate in all phases21.\n\nBy adopting a mixed methods design, employing qualitative and quantitative methods during consultation with mixed stakeholder groups, and by including a modified Nominal Group Technique component as described previously22, we aimed to provide a picture of the perceived benefits and challenges of Pulmonary Rehabilitation Programmes for COPD patients. Here we report the quantitative analysis of the Nominal Group Technique activities.\n\n\nMethods\n\nFollowing regional ethics and research and development approval, a series of consultation workshops were held between January and December 2012, in a District General Hospital in Wales, United Kingdom, serving a mixture of urban and agricultural communities. The hospital delivers a regular Pulmonary Rehabilitation Programme which includes 18 sessions of outpatient multidisciplinary input from occupational therapists, physiotherapists, dietetics staff, physicians, specialist respiratory nurses, social workers and a smoking cessation counsellor. The content and timings of the Pulmonary Rehabilitation Programme is evidenced-based and is tailored to individual requirements and personalised goal setting.\n\nWe recruited across the South West Wales Regional Health Board, United Kingdom that serves 385,000 people and included patient, professional and significant other groups, to ensure we included a wide range of views, experience and knowledge of COPD and Pulmonary Rehabilitation Programmes.\n\nPatients with COPD who were currently participating in or who had completed a Pulmonary Rehabilitation Programme within the last 2 years were approached to participate in the study, with most being approached in their last Pulmonary Rehabilitation Programme session. Information sheets were given to patients for their significant others (husbands, wives, partners, friends, carers or family members) inviting them to contact the researcher if they wished to participate. Professionals who were identified as playing a significant role in the delivery of the Pulmonary Rehabilitation Programmes and the treatment of COPD patients (occupational therapists, physiotherapists, respiratory consultants, respiratory team administrators, pharmacists, counsellors, psychologists, and specialist respiratory nurses) were also approached to participate in the study. All 20 participants (8 patients, 8 professionals and 4 significant others) provided written informed consent.\n\nOur aim was to gain an understanding of the positive and challenging aspects of Pulmonary Rehabilitation Programmes for patients with COPD and to gain a consensus regarding what constitute the most important aspects of Pulmonary Rehabilitation Programmes.\n\nNominal Group Technique consensus exercises were carried out as one aspect of a multi-layered, mixed-method consultation during three half-day workshops (one with professionals, one with COPD patients, and one with the significant others of patients). Based on guidance in the literature for optimal numbers for qualitative group consultations, we aimed to recruit six participants to each of the three workshops23.\n\nEach workshop was made up of three parts. Part one began with a broad discussion that examined the nature and content of Pulmonary Rehabilitation Programmes through a semi-structured group interview. The second part involved more extensive discussion with participants. Having attended a Pulmonary Rehabilitation Programme, participants were encouraged, using personal examples to describe what the Programme meant to them. This included exploring their perceived views regarding the benefits and challenges of Pulmonary Rehabilitation Programmes and impact on patient Health Related Quality of Life. An adapted Nominal Group Technique exercise was employed in the final part of the workshop. The focus of this stage was to address the following question with participants: “what are the positive, and what are the challenging aspects of Pulmonary Rehabilitation Programmes for the treatment and rehabilitation of COPD patients?” During the Nominal Group Technique exercise, issues that were raised in the early parts of the workshop were refined and condensed into a list of approximately ten positive and ten challenging aspects. At the end of the workshop, participants were asked to rank these aspects in order of significance (Steps 1–7, leading to Output 1, Figure 1). The generation of the positive and challenging aspects of the Pulmonary Rehabilitation Programme using Nominal Group Technique followed the standard approach outlined in previous work22.\n\nThe data generated from each Nominal Group Technique activity (Output 1, Figure 1) were collated for each consultation workshop. Median ranks with interquartile ranges were calculated using SPSS version 19 for each of the aspects on the positive and challenging lists and a consensus ranked list was produced based on these final median ranks.\n\nFollowing the consultation workshops we adapted the Nominal Group Technique method as previously described22 in order to include an additional multi-group ranking round (Steps 8–10, Figure 1). The lists of positive and challenging aspects of a Pulmonary Rehabilitation Programme produced following the three workshops were organised into a series of over-arching themes under which the positive and challenging aspects fitted (Step 8, Output 2, Figure 1). Rigour was maintained throughout the process of theme generation, by adhering to recommended qualitative data reliability and validity techniques24–26. An independent analysis of the lists generated from the workshops was carried out by two of the study team in order to identify the key over-arching themes. This process involved deletion of duplicate items and amalgamation of items where overlap was clear. A final set of common themes was independently generated by a third member of the team. This reflected and amalgamated the thematisations of the first two.\n\nFollowing the generation of themes, all the original workshop participants were sent a pack of A5-sized cards. Each card carried a broad theme as a header under which were listed the associated set of positive and challenging aspects. As with the earlier workshop Nominal Group Technique activity, participants were asked to rank the themes in order of importance: with ‘1’ representing the theme they regarded as being most important and subsequent ranks signifying the themes of diminishing importance (Step 9, Figure 1)22. The ranked cards were returned by participants in a pre-paid envelope.\n\nThe data from the returned cards were analysed using SPSS version 19 in order to calculate the median ranks and interquartile ranges (IQR) for each of the themes. A final consensus ranked thematic list was produced based on these median ranks (Step 10, Figure 1). This was the list produced for discussion and dissemination ensuring veracity within the method and enabling cross-consideration of themes and aspects by team members from Stage 1 thematisation undertaken within a group setting, to Stage 2 thematisation, undertaken by individual participants, post-consultation workshop.\n\nNotes and audio recordings from the three consultation workshops were transcribed. These transcripts were subjected to thematic and summative analysis to extract relevant information related to each of the generated themes27,28. The detailed content relating to each theme was extracted from the individual transcripts and was built up to articulate fully the set of aspects that it contained and to clarify any anomalies or ambiguities29. The final output of the consultation workshop was a ‘thematic template’ that ranked each theme in order and that provided a qualitative in-depth elaboration of the content contained within each theme.\n\n\nResults\n\nWe recruited a total of 20 participants across the three consultation workshops (see Table 1). Thirty three positive and 35 challenging aspects of Pulmonary Rehabilitation Programmes were produced in total for the three workshop group. The ranked list for each of the consultation workshops is illustrated in Table 2.\n\nSO, significant others; PROF, professionals; PT, patients; n/a, not available.\n\nThe positive and challenging aspects within each workshop group list represent the ranked lists ordered by the individuals in each group. The aspects generated are based on direct quotes from the individuals attending the workshops.\n\nIndividual assimilation produced similar lists of common broad themes that were refined to seven (Output 2, Figure 1). The seven themes were: the patient, physical health, mental health, knowledge and education, the programme, professionals and significant others and the future (see Table 3).\n\nThe positive and challenging aspects encompassed within each theme are based on direct quotes from the individuals attending the workshops.\n\nFourteen of the 20 attendees at the three workshops returned the packs of cards. Two were incorrectly completed, resulting in 12 evaluable responses (60%).\n\nFollowing thematic ranking, the theme that was regarded as most important was the patient, followed by physical health. Jointly ranked as third were: mental health and knowledge and education. The programme and professionals and significant others were jointly ranked as fifth, with the future ranked as the least important theme (Table 4).\n\n*Based on 12 evaluable responses. Total respondents=14 (14/20=70%), 2 were excluded from analysis due to incorrect completion.\n\nIn summary, the patient detailed how the patient’s health and wellbeing changed for the better over the course of Pulmonary Rehabilitation, and how patients were encouraged to gain confidence, to demonstrate a commitment to improving their own health, and to adopt a broader outlook on ongoing healthcare needs and expectations. Physical health illustrated how learning to breathe “properly” had a profound impact on patients, not only because breathing well is vitally important to their health and quality of life, but also because breathing “properly” is something that needs to be learnt. Mental health highlighted that bringing patients together enabled them to appreciate that they were not alone in their feelings and experiences. Knowledge and education emphasised the ability of Pulmonary Rehabilitation Programmes to create a learning environment, lasting for many weeks, within which patients are educated about their illness, and are able to develop new techniques to manage and cope. In the programme, patients, professionals, and significant others all emphasised positive outcomes for patients attending Pulmonary Rehabilitation Programmes for the duration and in the longer-term: physically, mentally, and socially. Professionals and significant others discussed how patients regarded the professionals as “caring” and “friendly”, treating them with “dignity” and “respect”, and that this created a welcoming and safe environment that enabled them to feel “cared for” and “at ease”. With respect to the theme of the future, participants emphasised a plethora of benefits that could be directly attributed to Pulmonary Rehabilitation Programmes, including improved health outcomes, enhanced quality of life, fewer hospital admissions, less time spent in hospital and consequently health care financial savings.\n\n\nDiscussion\n\nWe identified important aspects of Pulmonary Rehabilitation Programmes for the treatment of COPD from the point of view of a mixed population group of patients, professionals and significant others. Using a modified Nominal Group Technique exercise delivered during innovative consultation workshops, we produced a novel ranked thematic list that encompassed the important positive but also challenging aspects of Pulmonary Rehabilitation Programmes.\n\nThere was a surprisingly diverse range of generated aspects (Table 2) across the three workshops. The professional outputs were focused on pragmatic service delivery, with a clear goal of patient improvement, education and attitudinal change. The patients focused not only on physical improvements but also on improving mental strength, morale and self-esteem. Although all patients were positive about Pulmonary Rehabilitation Programmes, they also highlighted the challenges faced by some of them in attending them, which included an occasional lack of privacy, instances of poor communication, inadequate venues for certain activities (e.g. a public area of a hospital corridor to perform shuttle walk tests) and being daunted by the prospect of exercise and gym work. These findings are in accord with previous literature, which has examined the reasons for non-attendance on Pulmonary Rehabilitation Programmes13,17. Interestingly, the significant others focused on the social elements, with friendships made, caring staff and individual care contributing to the patients’ gaining confidence and learning about how to manage their condition. The significant others also highlighted the knock-on-effect of allowing them to have more time for themselves and not be so protective of the patients. All participants recognised that they were unsure what the future would bring in terms of long-term health and health-care support, but were keen for continued contact with professionals, Pulmonary Rehabilitation Programmes refresher courses and for the Pulmonary Rehabilitation Programmes to be recognised as beneficial for others, and thus maintained.\n\nThe final outcome of the Nominal Group Technique exercise was a ranked list of seven themes (Table 3), with ‘the patient’ ranked as the most important theme, followed by ‘physical health’. Overall, the main positive benefits of Pulmonary Rehabilitation Programmes were that they instilled confidence, enabled patients to breathe properly and manage their health more efficiently, encouraged the patient to be more self-sufficient and in control, and were enjoyable. The challenges to participation were that Pulmonary Rehabilitation Programmes were daunting, physically challenging, and required motivation. Interestingly, many of these challenges have been highlighted in previous qualitative studies16 with COPD patients as important reasons why patients decline entry or withdraw from Pulmonary Rehabilitation Programmes. Patient beliefs about Pulmonary Rehabilitation Programmes can comprise positive aspects (e.g. that they will lead to improvement, safe and multi-disciplinary setting, and motivation) as well as negative aspects (they lead to disruption of normal routine, being tired, transport issues and limited privacy)13. It has been shown that attending a Pulmonary Rehabilitation Programme is associated with better management of breathlessness, which in turn has a positive impact on physical and social activity, coping strategies and patient confidence15,17,18.\n\nThis study was carried out within one geographical location in South-West Wales, United Kingdom, and employed only three consultation workshops. Whilst we are confident that the methods adopted are transferable, in line with our extensive engagement with the methods in a range of community and primary care settings22,30, a larger study, employing more consultation workshops conducted over a larger geographical area is necessary to consider whether all the important aspects of Pulmonary Rehabilitation Programmes have been revealed, and whether the themes we identified within this study are generalisable.\n\nThe adapted Nominal Group Technique exercise was a mechanism for distilling the important aspects of Pulmonary Rehabilitation Programmes in a mixed group of individuals, which allowed the views of all the participating groups to be considered as equal. The process of qualitative elaboration of these themes in terms of what they meant to patients, professionals and significant others, provided a more comprehensive picture than other studies have derived. Moreover, combining qualitative with quantitative assessments provides more information, and these approaches could be used to make recommendations to improve and develop Pulmonary Rehabilitation Programmes across health-care contexts.\n\n\nData availability\n\nFigshare: Nominal Group Technique consultation of a Pulmonary Rehabilitation Programme Data Set, doi: 10.6084/m9.figshare.92854031\n\n\nConsent\n\nAll participants provided written informed consent.", "appendix": "Author contributions\n\n\n\nHH, FR, MD and SW developed the research proposal and applied for funding. HH and FR were joint principal investigators on the project. HH, FR, MD and SW were involved in the study consultation workshops. All authors were involved in the analysis and presentation of the study data. HH led on writing the manuscript. FR, MD, CC and SW provided input to and reviewed all drafts of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed. The study sponsor has had no involvement in the study design, collection, analysis or interpretation of data.\n\n\nGrant information\n\nThis work was supported by a grant from the Hywel Dda Local Health Board Research and Development Committee (Project R&D reference HD/12/004) and was awarded to Dr Hayley Hutchings and Professor Frances Rapport as joint principal investigators.\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 thank all of the study participants for giving their time and attending the workshops. We would also like to thank Carol-Anne Davies for her help in recruiting COPD patients and significant others, and Vicky Davies for her administrative support.\n\n\nReferences\n\nKelly C, Lynes D: Psychological effects of chronic lung disease. Nurs Times. 2008; 104(47): 82–85. PubMed Abstract\n\nStåhl E, Lindberg A, Jansson SA, et al.: Health-related quality of life is related to COPD disease severity. Health Qual Life Outcomes. 2005; 3: 56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHealthcare Commission. Clearing the air: a national study of chronic obstructive pulmonary disease. London: Commission for Healthcare Audit and Inspection. 2006. Reference Source\n\nStang P, Lydick E, Silberman C, et al.: The prevalence of COPD: using smoking rates to estimate disease frequency in the general population. Chest. 2000; 117(5 Suppl 2): 354S–9S. PubMed Abstract \| Publisher Full Text\n\nGlobal initiative for chronic obstructive lung disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. 2013. Reference Source\n\nNici L, Donner C, Wouters E, et al.: American Thoracic Society/European Respiratory Society statement on pulmonary rehabilitation. Am J Respir Crit Care Med. 2006; 173(12): 1390–413. PubMed Abstract \| Publisher Full Text\n\nRies AL: Pulmonary rehabilitation: summary of an evidence-based guideline. Respir Care. 2008; 53(9): 1203–7. PubMed Abstract\n\nNational Institute for Health and Clinical Excellence. Management of chronic obstructive pulmonary disease in adults in primary and secondary care. London, 2011. Reference Source\n\nLacasse Y, Goldstein R, Lasserson TJ, et al.: Pulmonary rehabilitation for chronic obstructive pulmonary disease. Cochrane Database Syst Rev. 2006; (4): CD003793. PubMed Abstract \| Publisher Full Text\n\nHill NS: Pulmonary rehabilitation. Proc Am Thorac Soc. 2006; 3(1): 66–74. PubMed Abstract \| Publisher Full Text\n\nBarnett M: Chronic obstructive pulmonary disease: a phenomenological study of patients' experiences. J Clin Nurs. 2005; 14(7): 805–12. PubMed Abstract \| Publisher Full Text\n\nRodgers S, Dyas J, Molyneux AW, et al.: Evaluation of the information needs of patients with chronic obstructive pulmonary disease following pulmonary rehabilitation: a focus group study. Chron Respir Dis. 2007; 4(4): 195–203. PubMed Abstract \| Publisher Full Text\n\nFischer MJ, Scharloo M, Abbink JJ, et al.: Participation and drop-out in pulmonary rehabilitation: a qualitative analysis of the patient's perspective. Clin Rehabil. 2007; 21(3): 212–21. PubMed Abstract \| Publisher Full Text\n\nKeating A, Lee A, Holland AE: What prevents people with chronic obstructive pulmonary disease from attending pulmonary rehabilitation? A systematic review. Chron Respir Dis. 2011; 8(2): 89–99. PubMed Abstract \| Publisher Full Text\n\nMonninkhof E, van der Aa M, van der Valk P, et al.: A qualitative evaluation of a comprehensive self-management programme for COPD patients: effectiveness from the patients' perspective. Patient Educ Couns. 2004; 55(2): 177–84. PubMed Abstract \| Publisher Full Text\n\nBulley C, Donaghy M, Howden S, et al.: A prospective qualitative exploration of views about attending pulmonary rehabilitation. Physiother Res Int. 2009; 14(3): 181–92. PubMed Abstract \| Publisher Full Text\n\nWilliams V, Bruton A, Ellis-Hill C, et al.: The effect of pulmonary rehabilitation on perceptions of breathlessness and activity in COPD patients: a qualitative study. Prim Care Respir J. 2010; 19(1): 45–51. PubMed Abstract \| Publisher Full Text\n\nCamp PG, Appleton J, Reid WD: Quality of life after pulmonary rehabilitation: assessing change using quantitative and qualitative methods. Phys Ther. 2000; 80(10): 986–95. PubMed Abstract\n\nJones J, Hunter D: Consensus methods for medical and health services research. BMJ. 1995; 311(7001): 376–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelbecq AL, van de Ven AH: A group process model for problem identification and program planning. Applied Behavioural Science. 1971; 7(4): 466–91. Publisher Full Text\n\nChapple M, Murphy R: The nominal group technique: Extending the evaluation of students' teaching and learning. Assessment and Evaluation in Higher Education. 1996; 21(2): 147–60. Publisher Full Text\n\nHutchings H, Rapport F, Wright S, et al.: Obtaining consensus about patient-centred professionalism in community nursing: nominal group work activity with professionals and the public. J Adv Nurs. 2012; 68(11): 2429–42. PubMed Abstract \| Publisher Full Text\n\nBloor M, Frankland J, Thomas M, et al.: Focus Groups in Social Research. London: Sage Publications Ltd, 2001. Reference Source\n\nSilverman D: Validity and credibility in qualitative research. In: Miller G, Dingwall R, editors. Context and method in qualitative research. The alternative paradigm. London: Sage, 1997; 12–25. Reference Source\n\nPope C, Ziebland S, Mays N: Qualitative research in health care. Analysing qualitative data. BMJ. 2000; 320(7227): 114–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeale C, Silverman D: Ensuring rigour in qualitative research. Eur J Public Health. 1997; 7(4): 379–84. Publisher Full Text\n\nBraun V, Clarke V: Using thematic analysis in psychology. Qual Res Psychol. 2006; 3(2): 77–101. Publisher Full Text\n\nRapport F: Summative Analysis: A qualitative method for social science and health research. Int J Qual Methods. 2010; 9(3): 270–90. Reference Source\n\nRyan GW, Bernard HR: Data management and analysis methods. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. 2nd ed. London: Sage, 2000; 769–803. Reference Source\n\nHutchings HA, Rapport FL, Wright S, et al.: Obtaining consensus regarding patient-centred professionalism in community pharmacy: nominal group work activity with professionals, stakeholders and members of the public. Int J Pharm Pract. 2010; 18(3): 149–58. PubMed Abstract \| Publisher Full Text\n\nHutchings HA, Rapport FL, Wright S, et al.: Nominal Group Technique consultation of a Pulmonary Rehabilitation Programme. Figshare. 2014. Data Source" }	[ { "id": "3648", "date": "05 Mar 2014", "name": "Kate Bullen", "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 and useful article that explores the implications and lived experience of people dealing with the challenges of a debilitating chronic disease (COPD). This is not a \"sexy\" area of health care research but it is an important one as COPD substantially limits life in both quantity and quality. As such it is a worthy area of investigation, as the aim of this article is to identify how to improve the delivery of service for a sometimes overlooked patient group. The article is well presented and reflects a high level of attention to detail in design and analysis. The research team has identified an appropriate method of investigation which is sufficiently novel to generate new knowledge that will potentially inform clinical practice. Details of the data collection and analysis processes are well presented and the data are discussed comprehensively. The research team acknowledges that there are limitations in the study in terms of the level of participants, and the level of geographical specificity of the study. Whilst this is a sensible approach such caveats should not detract from the potential value of the study for practitioners who deliver such programmers. Finally, the conclusions are well balanced and insightful with a clear pathway to future research identified.", "responses": [] }, { "id": "4411", "date": "08 Apr 2014", "name": "Pat G. Camp", "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\nSTUDY PURPOSE The purpose of the present study was to describe the perceived benefits and challenges of pulmonary rehabilitation program for patients with COPD from the point of view of patients, professionals, and significant others (carers). TITLE AND ABSTRACTThe title reflects the content of the article. The abstract provides a suitable summary of the work. The conclusion would be strengthened by a main message. INTRODUCTIONThe introduction provides detailed information about the area of research, gaps in the literature, and techniques employed to answer the research question. The references to the relevant literature on the mixed method design were also provided in order to support the proposed methodology. METHODSThe design of the study was also described clearly, including thematic generation and consensus, as well as responsibilities for integration. However, the authors state that the focus is on the quantitative analysis rather than the qualitative analysis, yet the analysis described appears to be qualitative in nature, and the authors comment on the need to maintain qualitative rigour throughout the study. A ranking by median scores by the participants is still qualitative in nature. If the intent is to undertake a quantitative analysis of the data, the selection of the median rank and interquartile ranges as appropriate statistical tests that is meant to answer the study research question seems insufficient. The authors might want to explore the differences between the different groups in regard to the perceived benefits and challenges of pulmonary rehabilitation program for patients with COPD. We consider that there is a need for clearly stating the hypothesis being tested and a motivation for selecting specific statistical analysis, should a quantitative analysis be intended. The individual data table would be better presented in the results section. RESULTSThe results section is appropriately explained, but the emphasis is again placed on the qualitative than quantitative research findings. In addition, the authors state that 20 participants were initially included in the study (8 patients, 8 professionals and 4 others) and 12 participants were finally included in the analysis. Given the small sample size, there is also a need to specify the status of the participants who did not complete the study. This information may influence the interpretation of the study results. There is also a disagreement between the evaluable responses across the paper: 60% on the page 6 versus 70% on the page 9. The authors also state that COPD patients who were currently participating in or completed a pulmonary rehabilitation program within the last 2 years were approached to participate in the study. However, Table 1 shows 2005 as year of PR program for one COPD patient. It is also worth mentioning the weak participation of physiotherapists in this study, although the pulmonary rehabilitation program was the main component of this research and the physiotherapists play a significant role in the delivery of pulmonary rehabilitation program. The thematic template generation does not offer clear guidance to the reader on what are the clear benefits and challenges – the list is somewhat self-evident and simplistic, and appears to focus on the benefits of the pulmonary rehabilitation program, but the challenges are not clearly articulated. DISCUSSIONThe conclusions are generally balanced and justified, although special emphasis is placed on the qualitative findings. Some of the discussion would be better suited to the results. The main benefits and challenges of pulmonary rehabilitation for patients, carers, and professionals as presented in the themes are not specific enough to be of use for making recommendations for improving the pulmonary rehabilitation program.", "responses": [ { "c_id": "915", "date": "22 Jul 2014", "name": "Hayley Hutchings", "role": "Author Response", "response": "We would like to thank the reviewers for their helpful comments. We have now made some changes to the original manuscript to help clarify some of the issues, or to highlight the limitations of the research. In addition we have responded to some of the specific queries below.The purpose of NGT is to employ quantitative and qualitative methods in order to gain a common consensus regarding the relative importance of generated issues. The findings are developed corroboratively using combined qualitative and quantitative approaches. The ‘quantitative’ analysis is a simple process whereby the most favoured rank is selected as being the most important. We have presented this prioritisation as the median rank with IQR as this is statistically more appropriate for small groups of data and better illustrates the range of ranks (i.e. the variability within the group) alongside the final priority list. We have followed the standard methodological approach for this analysis of ranking. NGT is not intended to be a specific statistical analysis method in this context, but rather a method for identifying broad priorities. The final prioritisation list produced may be subject to further debate, but the statistical approach used facilitates a starting point for further exploration. We extended the traditional NGT method by employing a further thematic stage which was designed to explore the more detailed rationale for the prioritised list generated. Although this has only been presented in simple terms in this paper, we generated an in depth template for each theme which elaborated upon the issues raised. It is beyond the scope of this paper to illustrate the in depth findings and these will be reported separately. The extended qualitative findings further explore the differences between groups in more detail, but we have tried to illustrate this briefly in this paper in Table 2 which lists the positive and challenging aspects across each of the groups.We feel that as the purpose of the paper is to present the consensus views rather than individual workshop findings, that these data tables are best appended in the data files rather than in the main results section. Everyone linked to the pulmonary rehabilitation programme in the two hospitals was invited to and participated in the workshops. It was our aim to gather views from those individuals running the programme at the Trust rather than physiotherapists who may have been involved in delivery of PR programmes elsewhere. We were not therefore in a position to invite more physiotherapists to the workshop as all those involved in delivery attended a workshop. The hospital ran a multi-disciplinary team approach to the delivery of the PR programme and we had representation from every aspect. No health professionals were excluded." } ] }, { "id": "5251", "date": "26 Jun 2014", "name": "Amanda Stears", "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 gives a clear introduction to the concept of Pulmonary Rehabilitation (PR) acknowledging the evidence supporting the intervention and the positive impact that is has on participants lives post programme. Most importantly, it seeks to gather the opinions of “significant others” and professionals within the clinical team, something that is often overlooked but that could be a key component in service development. There is a clear introduction to the use of the Nominal Group Technique and its usefulness in facilitating all participants’ views to be gathered and the aim of providing a rank of perceived benefits and challenges in pulmonary rehabilitation is clearly stated.\n\nThe authors have used the Nominal Group Technique successfully to provide a useful ranking of themes and challenges from the perspectives of patients, professionals and significant others. These are useful to challenge current assumptions within Pulmonary Rehabilitation Programmes, but the study would benefit from consideration of the socio-demographics of the group as these also impact on health related outcomes. The small sample size also makes it difficult to generalise to a wider population, something which the authors themselves acknowledge. There is also a strong male bias in the sample group, what affect did this have on the results? The data and techniques used are clearly described and well presented and are representative and reflective of findings in the clinical setting. In conclusion, going forwards further research is required to establish whether the rankings are concurrent across the country, and if so do they have a role in guiding PR and assisting patients and significant others in overcoming the perceived barriers. Many of the themes and issues identified are already known to those working within the clinical setting and are addressed through education and management within PR in the UK already. There are also assumptions in practice that patients “Learn” as is stated on page 6 in the Thematic Template generation. How is this evidenced? One finding that the authors also highlight is that significant others have more time for themselves as patients become more independent therefore supporting the self management of PR; this would benefit from further exploration.", "responses": [ { "c_id": "916", "date": "22 Jul 2014", "name": "Hayley Hutchings", "role": "Author Response", "response": "Thank you for these helpful comments. We have made some amendments to the manuscript on the basis of these comments which we hope will serve to clarify some of the issues raised." } ] } ]	1	https://f1000research.com/articles/3-42
https://f1000research.com/articles/3-185/v1	06 Aug 14	{ "type": "Software Tool Article", "title": "FALCON: a software package for analysis of nestedness in bipartite networks", "authors": [ "Stephen J. Beckett", "Chris A. Boulton", "Hywel T. P. Williams", "Chris A. Boulton", "Hywel T. P. Williams" ], "abstract": "Nestedness is a statistical measure used to interpret bipartite interaction data in several ecological and evolutionary contexts, e.g. biogeography (species-site relationships) and species interactions (plant-pollinator and host-parasite networks). Multiple methods have been used to evaluate nestedness, which differ in how the metrics for nestedness are determined. Furthermore, several different null models have been used to calculate statistical significance of nestedness scores. The profusion of measures and null models, many of which give conflicting results, is problematic for comparison of nestedness across different studies. We developed the FALCON software package to allow easy and efficient comparison of nestedness scores and statistical significances for a given input network, using a selection of the more popular measures and null models from the current literature. FALCON currently includes six measures and five null models for nestedness in binary networks, and two measures and four null models for nestedness in weighted networks. The FALCON software is designed to be efficient and easy to use. FALCON code is offered in three languages (R, MATLAB, Octave) and is designed to be modular and extensible, enabling users to easily expand its functionality by adding further measures and null models. FALCON provides a robust methodology for comparing the strength and significance of nestedness in a given bipartite network using multiple measures and null models. It includes an “adaptive ensemble” method to reduce undersampling of the null distribution when calculating statistical significance. It can work with binary or weighted input networks. FALCON is a response to the proliferation of different nestedness measures and associated null models in the literature. It allows easy and efficient calculation of nestedness scores and statistical significances using different methods, enabling comparison of results from different studies and thereby supporting theoretical study of the causes and implications of nestedness in different biological contexts.", "keywords": [ "Nestedness is a statistical property of systems where two kinds of entity interact", "which can be represented as bipartite networks. Originally used as a metric for species-site distributions1", "2", "nestedness has recently gathered much attention as a metric for bipartite species interaction networks", "e.g. plant-pollinator mutualisms3", "4 and host-virus interactions5–7. Various discussions have considered the sources of nestedness in such systems and its potential implications for ecological dynamics4", "8–13. However", "it is unclear how to systematically compare results for different ecological datasets. Furthermore", "nestedness is not restricted to ecological datasets", "but is a generic property of any bipartite network. Thus", "there is a need for measures of nestedness that are context-independent and do not depend on any particular (ecological) interpretation. Multiple methods for measuring nestedness have been used in different studies", "along with multiple approaches to calculating statistical significance of the measured values. This provides a large number of ways in which nestedness could be evaluated14–16. Before theoretical investigations of the mechanisms of nestedness can be properly undertaken", "robust measures and statistical tests for nestedness are required to allow comparison of results from different studies." ], "content": "Introduction\n\nNestedness is a statistical property of systems where two kinds of entity interact, which can be represented as bipartite networks. Originally used as a metric for species-site distributions1,2, nestedness has recently gathered much attention as a metric for bipartite species interaction networks, e.g. plant-pollinator mutualisms3,4 and host-virus interactions5–7. Various discussions have considered the sources of nestedness in such systems and its potential implications for ecological dynamics4,8–13. However, it is unclear how to systematically compare results for different ecological datasets. Furthermore, nestedness is not restricted to ecological datasets, but is a generic property of any bipartite network. Thus, there is a need for measures of nestedness that are context-independent and do not depend on any particular (ecological) interpretation. Multiple methods for measuring nestedness have been used in different studies, along with multiple approaches to calculating statistical significance of the measured values. This provides a large number of ways in which nestedness could be evaluated14–16. Before theoretical investigations of the mechanisms of nestedness can be properly undertaken, robust measures and statistical tests for nestedness are required to allow comparison of results from different studies.\n\nHere we present FALCON – a free software package that allows the user to easily compute several measures of nestedness and associated statistical significances based on a selection of null models. FALCON stands for “Framework for Adaptive ensembLes for the Comparison Of Nestedness”. FALCON operates on any form of bipartite interaction data represented as a matrix of associations and is set up to be deliberately ‘blind’ to the source and interpretation of input data. FALCON is based on the assumption that nestedness is a general statistical property of matrices and therefore its measurement should be independent of context or interpretation. FALCON calculates nestedness as a statistical property of a matrix, by returning the nestedness score for the most-nested configuration of the input matrix. Since calculating statistical significance of nestedness scores can be computationally demanding, involving generation of a large ensemble of matrices from a null distribution, FALCON uses a novel “adaptive ensemble” method to improve efficiency by using the minimal ensemble size sufficient to give robust statistics.\n\nSeveral software packages for calculating nestedness already exist – including1,11,17–21, but these are subject to various factors which make the direct comparison of different nestedness measures and the statistical interpretation of returned values difficult to achieve. Several nestedness measures are handled by packages which deliver a single measure, making the comparison difficult. Some are specific to a particular operating system. Some do not make the source codes available for re-implementation, reducing confidence in their outputs and prevent future extensions. Two packages for the R statistical programming language, bipartite19 and vegan21, together contain functions for several nestedness measures and associated null models, as well as many other tools for analysis of bipartite ecological networks. However, these packages offer no obvious implementation of significance testing (the principal method for reporting results of nestedness analyses) and they also lack several nestedness measures which have been recently developed. FALCON is designed to address these deficiencies, enabling the calculation of nestedness and statistical significance by using a variety of measures and null models, with open source code provided for several platforms.\n\nThe FALCON package is available for three commonly used numerical analysis platforms: MATLAB, Octave and R. MATLAB (http://www.mathworks.co.uk/products/matlab/) is a commercial software platform, while Octave (https://www.gnu.org/software/octave/) and R (http://www.r-project.org/) are both freely available open source platforms. FALCON can be freely downloaded on Github (http://github.com/sjbeckett/FALCON) or figshare22 and all code is open and accessible. A guide to downloading, installing and running FALCON accompanies the code. This document describes the assumptions on which FALCON is based, how it calculates nestedness and statistical significance, gives details of the adaptive ensemble method used to improve computational efficiency and provides a case study to demonstrate its usage and outputs.\n\n\n1 What is nestedness?\n\nNestedness is a statistical property of bipartite interaction data presented in matrix form. In a perfectly nested matrix, the entries in each successive row are a strict subset of those in the previous row, while the entries in each successive column are a strict subset of those in the previous column (Figure 1). Interpretation of nestedness depends on context.\n\nWhite squares indicate connections between two kinds of entity arranged in rows and columns.\n\nThe concept of nestedness was first described in studies on how species distributions varied between sites23–25, and later defined quantitatively as measuring the ‘amount of order/disorder’ in matrices representing the presence/absence of species in island communities1. Used in this way, nestedness is calculated from a matrix of presence-absence data where rows are species and columns are sampling sites along some environmental or spatial gradient. A perfectly “nested” matrix (see Figure 1) would be achieved when the set of species present at each site along the gradient is a subset of the species present at the previous site. Since then, the concept of nestedness has been extended in various directions; see26 for an historical overview of the nestedness concept. Nestedness has continued to be applied to spatial patterning (e.g.27) and has been linked with β-diversity28, but has also been applied to study mutualistic or antagonistic species-species interactions29,30, species-time relationships for a single site31, and several other types of bipartite networks9,10,32–35. For pairwise interactions (e.g. plant-pollinator or host-parasite systems), nestedness has been interpreted as placing species along a gradient of generalism-specialism in the number of partners they interact with; in this context, perfect nestedness is achieved when species within each class are ordered such that the interaction set (set of partners) for each species is a strict subset of that of the next species, and the most generalised species of one class interact with the most specialised species of the other class.\n\nNestedness is calculated from a biadjacency matrix representing pairwise interactions between two kinds of entity (one represented by rows, the other by columns). The order of rows and columns for a biadjacency matrix is arbitrary with respect to connectivity; rows and columns can be permuted without affecting the underlying topology of the interaction network. Any non-arbitrary ordering of rows and columns in the matrix representation necessitates the use of additional contextual information to specify which order rows and columns should take. While some datasets may suggest a “natural” ordering to rows and columns in the matrix representation of data (e.g. when one of the dimensions represents an environmental/spatial/temporal gradient), for many applications of nestedness there is no natural ordering (e.g. species interactions).\n\nAs stated above, we consider that nestedness should be a context-free metric, so that it can be applied to data without requiring any supplemental information on row/column ordering. This assumption implies that the ordering of rows and columns should not affect the measurement of nestedness. While some nestedness measures are insensitive to row/column ordering, several of the most commonly used measures are highly sensitive to ordering, introducing indeterminacy to the quantification of nestedness when rows/columns are ordered arbitrarily. To avoid this indeterminacy and return a single robust nestedness score for a given input matrix, FALCON can sort the rows and columns such that nestedness (however calculated) is maximised. Since re-ordering rows/columns in a matrix representation does not alter the structural information (node adjacency) of the underlying data, this re-ordering is a reasonable approach and makes the measurement of nestedness more consistent.\n\n\n2 Measures of nestedness in FALCON\n\nNestedness is most commonly calculated for binary data representing presence/absence of an interaction between two entities, but can also be calculated for weighted data that indicate the strength of the interaction. The methods used to calculate nestedness vary depending on whether binary or weighted interaction data are provided. The nestedness measures available in FALCON are shown and briefly described below and in Table 1; further details are given in Appendix A.\n\nThe nestedness measures considered here are not trivial variations upon each other, but differ significantly in their derivations. However, some similarities can be drawn. Spectral radius (SR)11 and the measure of Johnson, Domínguez-García, & Muñoz38 (JDM) are invariant to the ordering of rows and columns in the network and are calculated using the adjacency matrix of the network. On the other hand, discrepancy (BR)2, Manhattan distance (MD)37 and nestedness based on overlap and decreasing fill (NODF)36 are all sensitive to row/column ordering and are maximised when rows/columns are ranked by degree. The nestedness temperature calculator (NTC)1,21 involves sorting of rows and columns against the ‘isocline of perfect order’ (see Figure 5) such that it maximises connections above the isocline and minimises connections below the isocline. BR is similarly calculated relative to an idealised ‘maximally packed’ matrix. NODF is found through pairwise comparisons of overlap between subsequent rows and columns, whilst MD is found by assigning a weight to each connection as a sum of it’s row and column indexes. The measures also differ in how nestedness is scored; the degree of nestedness in a network increases with increasing measure score for JDM, NODF and SR, but with decreasing measure score for BR, MD and NTC.\n\n\n3 Comparison of nestedness scores\n\nNestedness is strongly sensitive to the size (number of rows and columns) and fill (number of non-zero entries) of the input matrix17. This is problematic in practical terms, since we often wish to compare nestedness of matrices that differ in these basic properties; in fact, cases where we compare empirically derived matrices with identical size and fill are an exception. Thus comparison of absolute values of nestedness metrics is not informative and may be misleading. To compare nestedness of matrices with differing size and fill, observed nestedness should always be interpreted in the context of a null distribution of matrices with similar properties. Measuring observed nestedness relative to expected nestedness derived from a null distribution of similar matrices allows determination of both effect size (e.g. as a z-score, which is commonly used to compare different nestedness schemes26,39) and statistical significance (e.g. as a p-value giving the expected frequency of the observed score in the null distribution). This approach necessitates choice of a suitable null model and generation of a distribution of random matrices drawn from it.\n\nIn the present context, a null model is a method for creating a distribution of matrices that conserves some properties of the input matrix while varying other properties at random40. We continue the “context-free” approach in our treatment of null models; to allow comparison of nestedness across different scenarios, a good null model should not make assumptions about the mechanisms by which data were generated, but treat the matrix as an independent data structure. However, to be comparable to the input matrix, null matrices must conserve some key matrix properties (such as size and fill) on which nestedness depends. The null models available in FALCON are given in Table 2; further detail is given in Appendix B. FALCON includes some of the more popular null models from the literature, alongside some additional null models that we feel can be useful. Null models vary in whether the original data is binary or quantitative, and in which properties of the original input matrix are preserved.\n\n\n4 How FALCON works\n\nFALCON requires several inputs:\n\nan input network in the form of a bipartite matrix\n\nwhether binary or quantitative nestedness should be investigated (quantitative matrices can be analysed using binary measures)\n\nwhether to sort rows and columns to maximise nestedness score\n\nwhich nestedness measures should be used\n\nwhich null models nestedness should be tested under\n\nwhether the ensemble of null models should be created with a fixed number or adaptively chosen\n\nwhether or not to plot the distributions of nestedness scores\n\nOutput is returned to the user in the form of:\n\nthe most nested configuration of the input matrix\n\nthe nestedness measure(s) of the input matrix\n\nthe expected value of nestedness under the null model(s) (as the mean measure of matrices created in the ensemble)\n\nthe number of ensemble members used to calculate significance in each null model\n\nthe statistical significance of the nestedness of the input matrix against each null model as a p-value\n\nthe standard deviation and sample z-scores of the measure in the ensemble as well as other properties.\n\nFALCON follows the process shown in Figure 2. First, it sorts the user input matrix into a maximally nested configuration and removes any empty rows/columns before finding the nestedness of this matrix using the users chosen measures. Then, FALCON goes through each of the user specified null models one by one, creating an ensemble of null matrices according to the rules of each null model. Each null matrix is then sorted and measured by each of the chosen nestedness measures. Thus, for each null model, nestedness measures are calculated for each of the null matrices in a single null ensemble, enabling direct comparison of results. The size of the null ensemble is determined by the input choice of using either the fixed or adaptive ensemble size (see Section 4.5). Statistics are computed from the measures found in the null ensemble (and the direction in which that nestedness measure is calculated), before the next null model ensemble is instantiated. Once all null models have been computed, the results are returned to the user.\n\nFor different nestedness measures, increasing scores can represent either increasing or decreasing nestedness as discussed in Section 2. FALCON initially determines whether a higher measure score is related to greater nestedness (or vice versa) in the chosen measure by comparing the scores returned for a highly nested network (see Figure 3A) and a highly non-nested network (a weighted checkerboard configuration; Figure 3B), for which the fill (number of non-zero elements) and element sums are equal. The direction of increasing nestedness for a given measure is used during calculation of statistical significance. This method of determining direction each time the algorithm runs is included to allow easy extensibility; if a new measure is added, FALCON will automatically determine which direction indicates increasing nestedness.\n\n(A) Weighted nested matrix. (B) Weighted non-nested matrix (checkerboard). Both (A) and (B) have 55 non-zero elements that sum to 220.\n\nFor efficiency, FALCON is set up to initially sort the input matrix by row and column degrees for calculation of BR, MD and NODF, retains this sorted configuration for calculation of JDM and SR, and subsequently re-sorts for NTC in order to find the maximal nestedness of a binary matrix. For quantitative data, FALCON uses the same methods as for binary interactions, but also utilises weight data to break symmetry when two rows (columns) have the same degree; in this case, the row (column) which has greater values for most overlapping elements is ranked highest. Where two or more rows (columns) share the same degree and most overlapping elements, the rows (columns) are ranked according to the total sum of row (column) elements. This sorting does not affect the underlying topology or the relationships in the data. FALCON also allows the user to decide if any sorting is performed, enabling the “context free” assumption to be relaxed (e.g. for investigation of gradient-based nestedness39).\n\nFALCON uses a bootstrap method to calculate the statistical significance of a given nestedness score, since the true null distributions of the test statistics are not known. The ensemble size used for this calculation can either be fixed or calculated adaptively by FALCON to improve computational efficiency and reduce undersampling effects. Note that the strongest significance that can be assigned is P<1N where N is the ensemble size.\n\nFixed. The number of null matrices used to make up the ensemble is fixed by the user. This method is effective providing that the ensemble is large enough to have statistical power; the larger the ensemble, the more power the test has and the closer the answer will be to the p-value for the (unknown) true null distribution. However, it is not obvious how large the ensemble needs to be; in the literature, amongst others30, use 1,000 null models in their ensembles, whilst12 use 10,000, and6 use 100,000. A large number of different null matrix configurations are possible for a given input matrix and we may wish to avoid undersampling42; however, at the same time very large ensembles can make the calculation of significance computationally intractable.\n\nAdaptive. FALCON includes a mechanism for adaptive determination of ensemble size. This is intended to ensure robust statistics are achieved, avoiding concerns about undersampling or oversampling42, while minimising computational load. The adaptive method works by creating two ensembles in parallel using the same null model. Starting with a minimum ensemble size of 500 in each group, the ensembles are expanded until they show similar statistical properties. This condition is met when the null hypothesis (both ensembles come from the same distribution) of a Mann-Whitney U-test cannot be rejected at 10% significance. When this occurs, it suggests each group represents a good sample of the underlying distribution, and the two groups are combined to form a single null ensemble used to calculate final statistics. The expansion of the size of the ensemble has an upper limit of 100,000 members in case the null hypothesis is always rejected. The adaptive ensemble methods balances statistical precision with computational efficiency; we conservatively use 1,000 as a minimum final ensemble size such that a p-value as low as 0.001 can be assigned.\n\np-value. The p-value is the probability that a matrix drawn from the null distribution will be more nested than the input matrix. Low values (p → 0) indicate that the input matrix is highly nested relative to the null distribution; commonly a threshold of p ≤ 0.05 or p ≤ 0.01 is used to denote a statistically significant level of nestedness. Here p is calculated by counting the frequency of matrices in the null ensemble that are more nested than the input matrix; for cases where no member of the null ensemble is more nested than the input matrix we conservatively assign P<1N where N is the ensemble size.\n\nNormalised Temperature. The normalised temperature is inspired by the τ-Temperature37. It describes the relationship between the nestedness measure found for the input matrix and the expected nestedness measure derived from the null model ensemble. It is described as:\n\n\n\nwhere < Measure > denotes the expected value. In simple terms, the normalised temperature indicates whether the input matrix is more or less nested than the expectation for a null distribution of similar matrices. Where the measure gives increasing scores with increasing nestedness, T > 1 indicates greater-than-expected nestedness. Where the measure gives decreasing scores with increasing nestedness, T < 1 indicates greater-than-expected nestedness.\n\nMean. The mean average of the set of nestedness measures found for each of the ensemble members is returned.\n\nStandard Deviation. The standard deviation (σ) of the set of nestedness measures found for each of the ensemble members is returned.\n\nSample z-score. The z-score, or standard score, is calculated as the difference between the nestedness measure and its expected value divided by the standard deviation of the sample:\n\n\n\nIt is a measure of the number of standard deviations the nestedness measure of the input matrix is above the expected value. Hence, the way it should be interpreted, as with the normalised temperature, depends on whether nestedness increases with increasing measure score.\n\n\n5 FALCON usage - case study\n\nTo demonstrate FALCON we analyse nestedness analysis in a bipartite network representing the hashtags used by a sample of Twitter users. Data were collected using the Twitter API (https://dev.twitter.com/docs/api) by searching for tweets including the hashtag “#IPCC” in the time period 21st September 2013 – 5th October 2013. A list of all hashtags used by all users found in the search dataset was then used to create a binary bipartite adjacency matrix for users and hashtags. This was then sampled to create a smaller matrix used for this case study by including each row/column with probability of 1.1 and removing any empty rows/columns. The resulting matrix was stored in a comma-separated file called ’IPCC_HTuse_10_10_1_53x27.csv’.\n\nThe box below shows the command sequence used to perform a binary nestedness analysis using FALCON in MATLAB. The first line reads in the “.csv” datafile, which includes row and column headers. The second line extracts the adjacency matrix from the imported data. The third line runs FALCON, using two binary nestedness measures (NODF and SR) and two null models (CC and FF, i.e. nulls 2 and 3) using the adaptive solver and displaying histogram plots. The fourth line plots the input matrix in its most nested configuration, as determined by FALCON. The nested configuration of the matrix and output histograms from significance testing are shown in Figure 4, whilst Table 3 shows an example output from the significance testing. Further examples for use of FALCON in R are given in supporting information accompanying the software.\n\n(A) the nested arrangement of the UserHashtag data. (B) Distribution of NODF scores found for an ensemble of FF null models generated for the UserHashtag network (C) distribution of spectral radius scores found from the same null matrix ensemble. The asterix marks the nestedness score of the input network.\n\nNestedness statistics were computed for the FF and CC null models using the NODF and spectral radius measures.\n\n\n\n\n6 Summary\n\nIn this paper we have presented FALCON, a software tool for reliable and efficient calculation of nestedness (and associated effect size and statistical significance) based on a selection of popular nestedness measures and null models used in the literature. FALCON treats nestedness purely as a statistical property of a bipartite matrix and removes any form of interpretation or contextual information from the analysis. This enables FALCON to be used to compare nestedness across a wide variety of application areas, noting that the concept of nestedness has already spread from its origin in island biogeography to include species-species interactions and other scenarios, and is likely to find further applications in other domains. The contribution of FALCON is to enable easy cross-comparison of observed nestedness using different nestedness measures and null models. We hope that this functionality will allow greater methodological uniformity and comparability of studies of nestedness. We are in the process of performing a large comparison study of nestedness metrics using FALCON (Beckett and Williams., in preparation). Uniformity of measurement and comparability of empirical results is an important preliminary step that must be achieved to enable understanding of the mechanistic basis and ecological (and otherwise) implications of nestedness. We hope that FALCON will be of use to other researchers and help illuminate this intriguing property of bipartite networks in many natural systems.\n\n\n7 Software availability\n\nThe FALCON software package is available from Github: http://github.com/sjbeckett/FALCON\n\nhttps://github.com/F1000Research/FALCON\n\nhttp://dx.doi.org/10.6084/m9.figshare.99911722", "appendix": "Author contributions\n\n\n\nSJB developed the initial project in MATLAB. SJB, CAB and HTPW enhanced the original package. SJB and CAB ported FALCON into R. SJB and HTPW wrote the manuscript with input from CAB.\n\n\nCompeting interests\n\n\n\nNo competing interests declared.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe thank Virginia Domínguez-García with the implementation of the JDM nestedness measure.\n\n\n8 Appendix A: Detailed description of nestedness measures in FALCON\n\nNODF. The nestedness measure based on overlap and decreasing fill (NODF) was first described by36 and has since become one of the most popular methods for describing the nestedness of a matrix. NODF can be found as:\n\n\n\nHere Ncol and Nrow are scores found by pairwise comparison of rows and columns, c is the number of columns, and r is the number of rows. Ncol is found as the sum of scores from pairwise comparisons of each column against all columns to its right. If both columns have the same degree, then the score is zero. If they have different degrees, the score is the percentage of elements in the second column which also appear in the first column. Nrows is found similarly for pairwise comparisons of each row against all rows below it. The sum of Ncol and Nrow is then normalised by the total number of pairwise comparisons. Values for NODF are between 0 (zero nestedness) and 100 (perfect nestedness). If the input matrix is first sorted to maximise nestedness by rank ordering rows and columns by degree, the form of NODF known as NODFMAX is found43.\n\nτ-Temperature and Manhattan distance. The τ-Temperature37 is a nestedness measure based on relative distances between matrix elements. Unlike other distance-based measures (such as NTC1 and its better described successors BINMATNEST17 and AININHADO18), the τ-Temperature does not use genetic algorithms to sort the data. The τ-Temperature is found by measuring the Manhattan distance D of the network matrix. This is the sum of the row and column indexes of all of the matrix elements Aij that are filled:\n\n\n\nManhattan distance is lower in more highly nested networks, since rows and columns can be shuffled so that many of the elements appear in upper-left positions where row and column indices are low. Once D is found, a null model is chosen (cf. Section 3) and an ensemble of null matrices are created. By finding the mean average Manhattan distance from the ensemble, denoted < Drand >, τ-Temperature can be calculated as:\n\n\n\nValues τ > 1 imply that D is greater than < Drand > and the network is less nested than expected for a network with the properties defined in the null model. τ is better described as a test statistic of the Manhattan distance, than as a measurement of nestedness itself.\n\nJDM Nestedness. The nestedness measure described in38, here termed JDM after author initials, treats nestedness as a measure of dissassortativity between the nodes, i.e., negative correlation between row and column degrees for non-zero elements of the input matrix. Their measure calculates the overlap (as the sum of the elements in the squared adjacency matrix which shows the minimum number of length two paths needed to connect any two nodes) of the input matrix and normalises it by the expected nestedness of the configuration model (a random graph with the same empirical degree distribution as the input network) and thus discounts the effect of degree heterogeneity. This nestedness score is unbounded, but when close to 1 it indicates that the matrix represents an uncorrelated random network. Unadjusted nestedness η˜ is calculated using the adjacency matrix a formed from the input bipartite matrix with r rows and c columns, where D is the node degrees in the adjacency matrix:\n\n\n\nNestedness of the configuration model ηconf can be calculated as:\n\n\n\nwhich can also be written as:\n\n\n\nwhere k are the row degrees and d are column degrees in the bipartite matrix. This leads to the normalised measure of nestedness for bipartite networks defined by38 as:\n\n\n\nNestedness Temperature. The original nestedness temperature calculator (NTC)1 was vaguely described and therefore difficult to reimplement, leading to several subsequent variations utilising similar underlying principles15,17,18,21. Here we have recoded the nestedtemp function from the R package vegan21. The nestedness temperature for an input matrix is based on the ‘isocline of perfect order’, a curve drawn from the lower-left corner of the matrix to the upper-right, with curvature defined by matrix fill (see Figure 5). Row and column orderings are then permuted using a genetic algorithm to maximise the number of connections above the isocline and minimise connections below the isocline. The number of connections which violate these rules, termed ‘surprises’, is then calculated and normalised to give a score between 0 (highly ordered) and 100 (highly disordered).\n\nHere there are two ‘surprises’.\n\nDiscrepancy. Discrepancy2, here denoted BR, quantifies nestedness as the difference between the input matrix and a perfectly nested matrix of the same dimensions and fill. The duplicate matrix P has the same row degrees as the input matrix, but the 1s in each row are pushed as far to the left as possible (ignoring the effect this has on underlying network topology). The discrepancy is then found by subtracting the input matrix from this perfectly nested matrix and counting the number of 1s that remain – the number of differences between P and the input matrix. By treating columns instead of rows an alternative perfectly nested comparator matrix P′ can be formed by pushing the 1s in each column of the input matrix as far to the top as possible – from which a different discrepancy score can be found. Here we modify the original method of2, which looks at discrepancy only in respect to P, and instead define discrepancy as the minimum of the individual discrepancy scores found from P and P′, to remove any bias towards row or column nodes.\n\nWNODF. The weighted NODF measure, WNODF20, uses a similar algorithm to NODF, but is designed for use on quantitative rather than binary networks. In addition to asking which pairs of rows/columns are subsets of one another, WNODF utilises weight information by also requiring that the preceding row/column has greater values in the overlapping elements. In effect, WNODF is a stricter version of NODF; the maximum WNODF score that can be achieved for a quantitative matrix is equal to the NODF score for the binary matrix.\n\nSpectral Radius. The spectral radius (SR) is defined as the absolute value of the maximum real eigenvalue from the adjacency matrix of a given input bipartite matrix. SR was proposed as a nestedness measure by11 and can be applied to both binary and quantitative matrices.\n\n\nReferences\n\nAtmar W, Patterson BD: The measure of order and disorder in the distribution of species in fragmented habitat. Oecologia. 1993; 96(3): 373–382. Publisher Full Text\n\nBrualdi RA, Sanderson JG: Nested species subsets, gaps, and discrepancy. Oecologia. 1999; 119(2): 256–264. Publisher Full Text\n\nBascompte J, Jordano P, Melián CJ, et al.: The nested assembly of plant-animal mutualistic networks. Proc Natl Acad Sci U S A. 2003; 100(16): 9383–9387. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJames A, Pitchford JW, Plank MJ: Disentangling nestedness from models of ecological complexity. Nature. 2012; 487(7406): 227–230. PubMed Abstract \| Publisher Full Text\n\nFlores CO, Meyer JR, Valverde S, et al.: Statistical structure of host-phage interactions. Proc Natl Acad Sci U S A. 2011; 108(28): E288–E297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFlores CO, Valverde S, Weitz JS: Multi-scale structure and geographic drivers of cross-infection within marine bacteria and phages. ISME J. 2013; 7(3): 520–532. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeckett SJ, Williams HT: Coevolutionary diversification creates nested-modular structure in phage-bacteria interaction networks. Interface Focus. 2013; 3(6): 20130033. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThébault E, Fontaine C: Stability of ecological communities and the architecture of mutualistic and trophic networks. Science. 2010; 329(5993): 853–856. PubMed Abstract \| Publisher Full Text\n\nSaavedra S, Stouffer DB, Uzzi B, et al.: Strong contributors to network persistence are the most vulnerable to extinction. Nature. 2011; 478(7368): 233–235. PubMed Abstract \| Publisher Full Text\n\nBustos S, Gomez C, Hausmann R, et al.: The dynamics of nestedness predicts the evolution of industrial ecosystems. PLoS One. 2012; 7(11): e49393. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStaniczenko PP, Kopp JC, Allesina S: The ghost of nestedness in ecological networks. Nat Commun. 2013; 4: 1391. PubMed Abstract \| Publisher Full Text\n\nMcQuaid CF, Britton NF: Host-parasite nestedness: A result of co-evolving trait-values. Ecol Complexity. 2013; 13: 53–59. Publisher Full Text\n\nLever JJ, van Nes EH, Scheffer M, et al.: The sudden collapse of pollinator communities. Ecol Lett. 2014; 17(3): 350–359. PubMed Abstract \| Publisher Full Text\n\nGotelli NJ: Null model analysis of species co-occurrence patterns. Ecology. 2000; 81(9): 2606–2621. Publisher Full Text\n\nUlrich W, Gotelli NJ: Null model analysis of species nestedness patterns. Ecology. 2007; 88(7): 1824–1831. PubMed Abstract \| Publisher Full Text\n\nCsermely P, London A, Wu L, et al.: Structure and dynamics of core/periphery networks. J Complex Networks. 2013; 1(2): 93–123. Publisher Full Text\n\nRodríguez-Gironés MA, Santamaría L: A new algorithm to calculate the nestedness temperature of presence–absence matrices. J Biogeogr. 2006; 33(5): 924–935. Publisher Full Text\n\nGuimarães PR, Guimarães P: Improving the analyses of nestedness for large sets of matrices. Environ Model Softw. 2006; 21(10): 1512–1513. Publisher Full Text\n\nDormann CF, Gruber B, Fruend J: Introducing the bipartite package: Analysing ecological networks. R News. 2008; 8(2): 8–11. Reference Source\n\nAlmeida-Neto M, Ulrich W: A straightforward computational approach for measuring nestedness using quantitative matrices. Environ Model Softw. 2011; 26(2): 173–178. Publisher Full Text\n\nOksanen J, Guillaume Blanchet F, Kindt R, et al.: vegan: Community Ecology Package. R package version 2.0–10. 2013. Reference Source\n\nBeckett SJ, Boulton CA, Williams HT: FALCON: nestedness statistics for bipartite networks. 2014. Data Source\n\nHultén E: Outline of the history of arctic and boreal biota during the Quaternary period: their evolution during and after the glacial period as indicated by the equiformal progressive areas of present plant species. Bokförlags Aktiebolaget Thule. 1937; 1: 168. Reference Source\n\nDarlington PJ: Zoogeography: the geographical distribution of animals. Wiley, New York, 1957; 676. Reference Source\n\nDaubenmire R: Floristic plant geography of eastern Washington and northern idaho. J Biogeogr. 1975; 2(1): 1–18. Publisher Full Text\n\nUlrich W, Almeida-Neto M, Gotelli NJ: A consumer’s guide to nestedness analysis. Oikos. 2009; 118(1): 3–17. Publisher Full Text\n\nFraser CI, Terauds A, Smellie J, et al.: Geothermal activity helps life survive glacial cycles. Proc Natl Acad Sci U S A. 2014; 111(15): 5634–5639. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaselga A: The relationship between species replacement, dissimilarity derived from nestedness, and nestedness. Glob Ecol Biogeogr. 2012; 21(12): 1223–1232. Publisher Full Text\n\nFortuna MA, Stouffer DB, Olesen JM, et al.: Nestedness versus modularity in ecological networks: two sides of the same coin? J Anim Ecol. 2010; 79(4): 811–817. PubMed Abstract \| Publisher Full Text\n\nJoppa LN, Montoya JM, Solé R, et al.: On nestedness in ecological networks. Evol Ecol Res. 2010; 12: 35–46. Reference Source\n\nChávez VA, Doncaster CP, Dearing JA, et al.: Detecting regime shifts in artificial ecosystems. In Advances in Artificial Life. ECAL. 2013; 12: 625–632. Publisher Full Text\n\nTim Tinker M, Guimarães PR Jr, Novak M, et al.: Structure and mechanism of diet specialisation: testing models of individual variation in resource use with sea otters. Ecol Lett. 2012; 15(5): 475–483. PubMed Abstract \| Publisher Full Text\n\nDalsgaard B, Trøjelsgaard K, González AMM, et al.: Historical climate-change influences modularity and nestedness of pollination networks. Ecography. 2013; 36(12): 1331–1340. Publisher Full Text\n\nPiepenbrink A, Gaur AS: Methodological advances in the analysis of bipartite networks: an illustration using board interlocks in Indian firms. Organ Res Meth. 2013; 16(3): 474–496. Reference Source\n\nMelo AS, Cianciaruso MV, Almeida-Neto M: treenodf: nestedness to phylogenetic, functional and other tree-based diversity metrics. Methods Ecol Evol. 2014; 5(6): 563–572. Publisher Full Text\n\nAlmeida-Neto M, Guimaraes P, Guimarães PR, et al.: A consistent metric for nestedness analysis in ecological systems: reconciling concept and measurement. Oikos. 2008; 117(8): 1227–1239. Publisher Full Text\n\nCorso G, Britton NF: Nestedness and τ-temperature in ecological networks. Ecol Compl. 2012; 11: 137–143. Publisher Full Text\n\nJohnson S, Domínguez-García V, Muñoz MA: Factors determining nestedness in complex networks. PLoS One. 2013; 8(9): e74025. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUlrich W, Almeida-Neto M: On the meanings of nestedness: back to the basics. Ecography. 2012; 35(10): 865–871. Publisher Full Text\n\nGotelli NJ: Research frontiers in null model analysis. Glob Ecol Biogeogr. 2001; 10(4): 337–343. Publisher Full Text\n\nStrona G, Nappo D, Boccacci F, et al.: A fast and unbiased procedure to randomize ecological binary matrices with fixed row and column totals. Nat Commun. 2014; 5: 4114. PubMed Abstract \| Publisher Full Text\n\nPoisot T, Gravel D: When is an ecological network complex? Connectance drives degree distribution and emerging network properties. PeerJ. 2014; 2: e251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPodani J, Schmera D: A comparative evaluation of pairwise nestedness measures. Ecography. 2012; 35(10): 889–900. Publisher Full Text" }	[ { "id": "5722", "date": "12 Aug 2014", "name": "C. Patrick Doncaster", "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 introduces a new freeware package designed to collate existing measures of nestedness in biological samples. This is a useful capability because several different methods exist for quantifying parameters and testing significance. The paper describes a novel and efficient method for sampling the null distributions that calibrate the significance of nestedness scores. The package should interest anyone working with biological networks of bipartite data, and is conveniently available for open-source platforms R and Octave as well as MATLAB. I have no major concerns about the conception, execution or description of the paper. It appears to me well motivated, technically sound and written with considerable clarity. The case study provides a useful illustration, though is under-developed as I will detail below and perhaps represents a missed opportunity. I found the program itself easy to use and really useful for exploring alternative measures of nestedness The manuscript title answers to the content of the paper. The Abstract provides a clear summary, though its impact would benefit from removal of repetition. An abstract of 150-words in a single paragraph could state concisely what the need is, and how the package addresses it. The Introduction could usefully explain the meaning of nestedness (currently in the following section) before briefly reviewing alternative methods of quantification. That review should attempt some categorisation of existing methods in one paragraph, and examples of their applications, before describing their various limitations in a following paragraph. Then say how FALCON addresses these issues. The paper provides a clear description of the concept of nestedness and its interpretation, which is not an easy task. It would help to define ‘bipartite’ in the context of the Fig.-1 matrices, as the two dimensions that make up the matrix columns and rows. Table 1 gives shorthand codes for the various nestedness measures, though only some of these correspond to the codes for calling the measures in the program (Table 2 of the instruction guide). It would help to have a closer correspondence of shorthand with code of measures, and with the subtitles in Appendix A. Likewise for null models and Appendix B. The Section-5 case study should take the opportunity to show readers how they can find interesting patterns in datasets by calculating nestedness, using sensible measures and calibrations. For this particular example, it would help to clarify the context by indicating that ‘IPCC’ stands for either ‘Intergovernmental Panel on Climate Change’ or ‘Independent Police Complaints Commission’. It’s not clear to me whether this hashtag was chosen purposefully to sample such different populations of Twitter users, and if so how one might interpret nestedness in the aggregate population. Any case study serves its purpose only if it is followed through all the way from motivation through to extraction of qualitative meaning from the quantitative analysis. Thus the text needs to interpret Figure 4 qualitatively, explaining the choice of measure and null model for these data, and resulting nestedness score such that the reader can see how nestedness might reveal interesting pattern in the data. It would be good to have the case-study dataset made available for users to try out for themselves. In the case study, the textual explanation of the command box says that the program uses three null models, but as far as I can see it calls only two: FF and CC. Then, confusingly the output graphs of null distributions show only the first one: CREATEBINNULL2. Actually, the second one looks more interesting according to Table 3. These steps from input to output need more explanation. Particularly for Fig. 4 and Table 3, interpretation is not helped by the lack of correspondence between names used for shorthand (e.g., ‘SR’, ‘FF’) and code (e.g., ‘SPECTRAL_RADIUS’, ‘CREATEBINNULL2’). If correspondence is not possible in the program itself, it would help at least to have clearer titles for graphs and tables. When using the package, I thought that the instruction guide would benefit from some tidying up. For example, it would help to tabulate all possible alternative values for each of the seven options. Null models should be presented after measures, since a null only has meaning as the baseline against which to calibrate a measure. For the R version of the program, an example script is provided to illustrate capability (examplescript.R). Its ‘TEST 2’ loads a .csv file containing a matrix suitable for nestedness analysis, but omits the specification ‘header = FALSE’. As a result, it reads the first line of the file as a header, which was surely not the intention. Specific points that merit attention: Last paragraph of Section 1, say: “FALCON can sort rows and columns so as to maximise nestedness regardless of the method used for its calculation.”When referring to specific published texts, remove the superscript, to say for example (Section 4.5): “in the literature, ref 30 amongst others use 1000 null models in their ensembles, whilst ref 12 use…” etc.Figure 4 component parts need labels (A), (B), (C). Label the axes of the matrix, at least in the legend if not in the figure itself: presumably columns for users and rows for hashtags.Table 3, avoid vertical lines in tables. The table would be easier to read if columns and rows were switched around. Left-justify text, and use a decimal tab to align numbers in a column.Avoid the phrase ‘As stated above’ – reorganise to state only once.Avoid ending a sentence in a verb. The verb imparts meaning to the sentence by describing the action, so giving it early facilitates the reader’s comprehension.", "responses": [] }, { "id": "6467", "date": "19 Nov 2014", "name": "Justin R. Meyer", "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 great, easy-to-use, and comprehensive piece of software. I am eager to begin to use it myself and to send it to colleagues who are reluctant to run the other nestedness software. The paper is also very well-written making this field more accessible. As the authors state, many forms of biological data are organized in bipartite networks and show nested pattern. With such a software platform and clear manuscript, I hope more similarities across biological systems will be recognized and appreciated, perhaps even promoting collaborations and new theory. My one suggestion is for the authors to provide a more thorough description of what null models should be used with which data structures, sampling procedures, biological processes, ecs. This would aid the cross-system comparisons and users from finding spurious patterns.", "responses": [] } ]	1	https://f1000research.com/articles/3-185
https://f1000research.com/articles/3-37/v1	05 Feb 14	{ "type": "Opinion Article", "title": "The second touch hypothesis: T cell activation, homing and polarization", "authors": [ "Klaus Ley" ], "abstract": "The second touch hypothesis states that T cell activation, proliferation, induction of homing receptors and polarization are distinguishable and, at least in part, sequential. The second touch hypothesis maintains that full T cell polarization requires T cell interaction with antigen-presenting cells (DCs, macrophages, B cells and certain activated stromal cells) in the non-lymphoid tissue where the antigen resides. Upon initial antigen encounter in peripheral lymph nodes (PLN), T cells become activated, proliferate and express homing receptors that enable them to recirculate to the (inflamed) tissue that contains the antigen. Differentiation into the T helper lineages Th1, Th2, Th17 and induced regulatory T cells (iTreg) requires additional antigen presentation by tissue macrophages and other antigen presenting cells (APCs) in the inflamed tissue. Here, I present a conceptual framework for the importance of peripheral (non-lymphoid) antigen presentation to antigen-experienced T cells.", "keywords": [ "Naïve αβ T cells are activated by contacting antigen-presenting cells in secondary lymphoid organs1. Dendritic cells (DCs) are the most effective antigen-presenting cells for naïve T cells. T cells must receive two signals", "one from T cell receptor (TCR) engagement with antigenic peptide in the context of MHC-II for CD4 or MHC-I for CD8 T cells", "and the other one through CD28 engagement by CD80 and/or CD862. Both the MHC and co-stimulatory signals are provided by the same dendritic cell3. The cytokine milieu at the time of antigen presentation drives T cell polarization. Here", "I will only consider helper (CD4) T cells", "but similar lineages", "subsets and polarizations have also been described for CD8 T cells." ], "content": "Traditional view\n\nNaïve αβ T cells are activated by contacting antigen-presenting cells in secondary lymphoid organs1. Dendritic cells (DCs) are the most effective antigen-presenting cells for naïve T cells. T cells must receive two signals, one from T cell receptor (TCR) engagement with antigenic peptide in the context of MHC-II for CD4 or MHC-I for CD8 T cells, and the other one through CD28 engagement by CD80 and/or CD862. Both the MHC and co-stimulatory signals are provided by the same dendritic cell3. The cytokine milieu at the time of antigen presentation drives T cell polarization. Here, I will only consider helper (CD4) T cells, but similar lineages, subsets and polarizations have also been described for CD8 T cells.\n\nTh1 develop in response to IL-12 and IFN-γ, express the defining transcription factor T-bet (Tbx21) and secrete the signature cytokine IFN-γ\n\nTh2 develop in response to IL-4, IL-5 and IL-13, express the defining transcription factor GATA-3 (Gata3) and secrete the signature cytokines IL-4, IL-5 and IL-13\n\niTreg develop in response to TGF-β, express the defining transcription factor FOXP3 (Foxp3) and secrete the signature cytokine IL-10\n\nTh17 develop in response to IL-6, TGF-β and IL-1, express the defining transcription factor ROR-γt (Rorc) and secrete the signature cytokines IL-17A, IL-17F and IL-21\n\nFollicular helper T cells (TFH) develop in response to CD40 and ICOS ligand, express the defining transcription factor BCL6 (Bcl6) and secrete the signature cytokine IL-21\n\nIn the traditional view, naïve T cell activation, proliferation and differentiation are all considered as concomitant, simultaneous processes.\n\n\nAntigen transport to lymph nodes\n\nAntigenic proteins reach draining lymph nodes either as soluble antigens or by way of migratory DCs that acquired antigen in the inflamed tissue. Soluble antigen is presented by at least 3 types of lymph node-resident DCs to naïve T cells. Multiple short (3 minutes) interactions allow the T cell to become activated. A little later, migratory DCs enter the draining lymph node through afferent lymphatics and present higher doses of antigen, or more highly processed antigen, to naïve T cells, resulting in long-term (~50 min) interactions2. Whether these longer-term interactions activate T cells more fully than repeated 3 min interactions is unclear. Many, if not all, migratory DCs are monocyte-derived, and only migrate to the regional lymph nodes under conditions of inflammation. These DCs must undergo a “maturation”, which involves expression of the chemokine receptor CCR7, and upregulation of the co-stimulatory molecule CD86. Maturation can be achieved by TLR ligands such as LPS or by engagement of CD40. Only mature DCs acquire the ability to migrate to lymph nodes4. Mature DCs share phenotypic characteristics and gene expression with M1 macrophages. M1 macrophages are characterized by the expression of iNOS, producing nitric oxide (NO) from arginine5. M2 macrophages express arginase, an enzyme that converts arginine to ornithine. The environment (M1, M2) during antigen encounter is likely to influence the ensuing T cell response.\n\n\nT cell activation\n\nNaïve T cell activation in peripheral lymph nodes is a dramatic and violent process6. Most activated T cells die by apoptosis in the lymph node where they were generated. The resulting apoptotic bodies are phagocytosed by local macrophages, triggering an anti-inflammatory signal7,8 that induces iTregs9. iTregs are a natural by-product of any ongoing immune response, whether in infections or autoimmune diseases10. Some activated T cells survive and acquire activation markers including:\n\nExpression of CD69 (which inhibits the sphingosine-1-phosphate receptor S1P111 and helps to keep the activated T cell from recirculating)\n\nExpression of CD25 (a subunit of the IL-2 receptor)\n\nProduction and secretion of IL-2\n\nExpression of CD44, a hyaluronic acid receptor\n\nDownregulation of CD62L (L-selectin)\n\nAntigen-experienced T cells proliferate in the lymph node for 3–7 days and acquire peripheral homing receptors that allow them to return to the tissue where the antigen resides.\n\n\nHoming receptors\n\nHoming receptors are defined as cell surface molecules that allow T cells to attach to endothelial cells in specific tissues and organs, transmigrate through the endothelial cell monolayer and the basement membrane and infiltrate the inflamed tissues. One category of homing receptors are chemokine receptors; heptahelical cell surface receptors that are coupled through heterotrimeric G-proteins. These include:\n\nCXCR3 and CCR5 in Th1 cells\n\nCCR4 in Th2 cells\n\nCCR6 and CXCR6 in Th17 cells\n\nCCR9 and 10 for T cells homing to the small and large intestine, respectively\n\nCCR7 for naïve T cells and central memory T cells; both home to lymph nodes\n\nOther homing receptors are adhesion molecules.\n\nAntigen-experienced CD4 T cells can be induced to express α4β7 integrin through a retinoic acid-dependent mechanism12.\n\nPSGL-1, a scaffolding glycoprotein that carries ligands for selectins. PSGL-1 is not functional on naïve T cells, but becomes functional upon induction of fucosyl transferase-VII; an enzyme highly expressed in Th1 cells13.\n\nCutaneous lymphocyte antigen (CLA) is a collection of glycoproteins that can serve as E-selectin ligands and drive homing of Th2 cells to the skin\n\nL-selectin (CD62L) is a homing receptor for naïve T cells and central memory T cells to home to lymph nodes\n\nThe Th1 chemokine receptor CXCR3 is induced soon after a naïve T cell sees antigen14. Interestingly, induction of most homing receptors is T-bet dependent15.\n\n\nT cell polarization\n\nCD4 T cells differentiate to Th1, Th2, Th17, iTreg and TFH cells. All CD4 T cell expansion and activation is IL-2-dependent. It is useful to emphasize that IL-2 deficient mice suffer from autoimmune disease and not from immunodeficiency. The autoimmune phenotype is driven by a lack of regulatory T cells. This suggests that the primary function of T cell activation and expansion is the production of iTregs, which appears to be necessary because central tolerance (elimination of self-reactive T cell clones in the thymus) is ineffective16,17. Although T cell clones are eliminated in the thymus by negative selection, other T cell clones differentiate to natural regulatory T cells (nTregs). nTregs show chromatin remodeling that allows stable FoxP3 expression. nTregs are not produced in secondary lymphoid organs and thus are not subject to the second touch hypothesis and therefore not considered further.\n\nThese induced regulatory T cells are mainly dependent on TGF-β1, a product of M2 macrophages and many tissue cells18. The dominant role of TGF-β1 is demonstrated by lethal autoimmune disease in TGF-β1 knockout mice19, which is phenocopied by T cell-specific disruption of TGF-β receptor II20. iTregs express the transcription factor FoxP3. FoxP3 expression is induced by IL-27 and IFNγ through a STAT1-dependent mechanism, by TGF-β through SMAD3, TIEG and ITCH, and by retinoic acid (RA) through an unknown mechanism10,21. FoxP3 is inhibited by IL-27 through a STAT3-dependent mechanism, by IFN-γ through IRF1, by IL-4 through STAT6, by IL-5 or IL-6 through STAT5 and by S1P1, CD28 and IL-21 through unknown mechanisms. Under the influence of inflammatory cytokines like IL-6, iTregs can differentiate to Th1722.\n\nThese IL-17A- and IL-17F-producing CD4 T cells play central roles in host defense against fungal and some bacterial infections, but also in many autoimmune diseases. The defining transcription factor of Th17 cells is RORγt (Rorc), but they also express RORα, IL-21 and IL-23R. Th17 cells are induced by IL-6 or IL-1 and TGF-β, but also IL-9 and IL-21 through a STAT3-dependent mechanism. Th17 cells are induced by segmented filamentous bacteria in the intestinal flora and by DCs activated by dectin-1 ligands. Th17 differentiation is inhibited by IL-27 (through STAT1) and by Th1- and Th2-driving cytokines. Some Th17 cells also express IL-1023.\n\nThese T-helper 1 cells secrete interferon-γ and TNF. Th1 differentiation is promoted by IL-12 through a STAT4-dependent signaling pathway and by IFN-γ through STAT1. The defining transcription factor is T-bet, encoded by the Tbx21 gene. T-bet inhibits other directions of CD4 T cell differentiation.\n\nT-helper 2 cells produce IL-4, IL-5 and IL-13. Although Th2 cells can develop in response to IL-4, IL-5 and IL-13, the Th2-inducing stimulus from antigen-presenting cells is not known. The defining transcription factor is GATA-3, encoded by the Gata3 gene.\n\nFollicular helper T cells develop in response to ICOS ligand and CD40. They migrate to B cell areas and provide essential help to B cells as they undergo differentiation and switch the antibody isotypes they secrete. The defining transcription factor is Bcl6.\n\nCD4 T cells produce memory cells that can be divided into effector- (TEM), central memory- (TCM)24, and into resident memory cells (TRM) that stay in the skin and intestinal epithelium25. TEM are CD45RBlow, CD44high and LFA-1high. TCM express CCR7 and CD62L. TRM express CD103 and CD69. It is currently controversial whether long-lived memory T cells originate from effector T cells (the cells expanding during the acute insult) or represent a separate lineage26. There is some evidence that memory T cells retain memory of the phenotype (Th1, Th2, Th17) they had during the initial encounter of antigen.\n\n\nThe role of innate immune cells\n\nInnate immune cells drive the polarization of the adaptive immune response. This was first shown in mice: certain inbred mouse strains have a Th1 bias (like C57BL/6) and others have a Th2 bias (like Balb/c). The underlying reason for this bias is not in the T cell compartment, but is found in the innate immune system5. This was shown by using immunodeficient mice that do not have an adaptive immune system, yet still preserve the Th1/Th2 bias. This finding gave rise to the concept of M1 and M2 macrophages5, where M1 macrophages metabolize arginine to NO and citrulline through iNOS and M2 macrophages metabolize arginine to ornithine and urea through arginase. NO promotes bacterial killing and inflammation, whereas ornithine promotes wound healing and collagen biosynthesis.\n\nM2 macrophages produce TGF-β, which is secreted in an inactive form that requires activation by a process that depends on αVβ8 integrin27–29. The activating integrin must be expressed on the same cell that presents the antigen, suggesting that only TGF-β secreted by αVβ8 integrin-expressing macrophages and DCs is relevant for iTreg and Th17 induction.\n\n\nMacrophage activation\n\nMacrophages express a large number of cell surface and cytosolic receptors that allow them to recognize bacteria, fungi, parasites, viruses, but also altered self molecules. The macrophage activating receptors fall into five families:\n\nToll-like receptors (TLRs)\n\nNOD-like receptors (NLRs), which are central to the assembly of the inflammasome and production of the IL-1 and IL-18 family of inflammatory cytokines\n\nRIG-I like receptors (RLRs)\n\nC-type lectins, which includes receptors for fungal pathogens like dectin-1\n\nScavenger receptors, including scavenger receptor-A, B and CD36\n\nEngagement of many of these receptors activates members of the transcription factor family NF-kB, which in turn is responsible for production of many inflammatory cytokines. How the engagement of specific pattern recognition receptors is related to the cytokines that drive CD4 T cell differentiation is an area of active investigation. Macrophages can secrete IL-1α and β, IL-18, IL-6, IL-12, IL-23, α and β interferons, IL-4, IL-10 and TGF-β. Activated macrophages express a range of molecules from the TNF and TNF receptor superfamilies.\n\n\nSecond touch hypothesis\n\nThe second touch hypothesis as formulated here, states that T cell activation, proliferation, induction of homing receptors, and polarization are distinguishable and at least in part sequential. The second touch hypothesis maintains that full CD4 T cell polarization requires CD4 T cell interaction with antigen-presenting cells (DCs, macrophages, B cells and certain activated stromal cells) in the inflamed non-lymphoid tissue where the antigen was first encountered (Figure 1).\n\nNaïve T cells (light blue) enter peripheral lymph nodes (PLN) through high endothelial venules (HEV), where they encounter antigen (Ag) presented by a migratory dendritic cell (DC) or by a PLN-resident B cell (brown). If signals induce Bcl6 and CXCR5, the T cells may enter the germinal center and become follicular helper T cells (TFH, red). Upon encountering antigen in the context of co-stimulatory molecules like CD80 and CD86, most T cells will express homing receptors, downregulate the sphingosine-1-phosphate receptor S1P1 and leave the PLN. These cells (orange) are only partially programmed and referred to as pre-Th1, pre-Th2, pre-Th17 and pre-iTreg. These cells circulate and reach various tissues, including the inflamed tissue from which the antigen came. There, they encounter DCs, macrophages and other antigen-presenting cells (APCs) such as endothelial cells. These cells present antigen in the context of different co-stimulatory molecules such as TNFSF members, and in the context of inflammatory cytokines found in the inflamed tissue. If the prevailing signal is IL-12, the T cell will commit to Th1, if IL-4, 5 and 13, to Th2, if IL-1, TGF-β, IL-6, IL-21, to Th17. If the cytokine environment is dominated by TGF-β, an M2 cytokine, the T cell will become an iTreg.\n\nThis hypothesis holds that homing receptors are induced first, soon after activation of naïve T cells. Expression of homing receptors capacitates these T cells to leave the lymphatic system and enter specific tissues, including those where the antigen was first encountered.\n\nThe second touch hypothesis also holds that antigen-experienced T cells require a “second touch” by seeing antigen presented on a tissue macrophage or DC in the context of co-stimulatory molecules and in the inflammatory cytokine environment of the inflamed tissue (Figure 2). This “second touch” appears to be required for full differentiation to Th1, Th2, Th17, Treg or TFH. This may involve chromatin remodeling of the loci needed for the respective differentiation (see below).\n\nTissue macrophages have a natural bias toward M2 (express arginase, produce TGF-β). T cells in this environment are likely to commit to the iTreg lineage. If tissue macrophages sense the presence of pathogens through their TLRs, NLRs, RLRs (see definitions in the main text) or C-type lectin receptors, they convert to M1 (express iNOS, produce IL-12, IL-23, upregulate MHC-II and CD86). M1 differentiation also induces expression of chemokines that attract more monocytes from the circulation, which have a propensity to become M1-polarized after transmigration. T cells receiving a second touch in this environment are likely to commit to Th1 if IL-12 dominates or Th17 if IL-23 dominates. The tissue environment producing Th2 cells is not well understood, but may involve M2 macrophages. It has also been proposed that Th2 cells arise when basophils present antigen51.\n\n\nAntigen presentation in non-lymphoid tissues\n\nAntigen presentation to antigen-experienced T cells in non-lymphoid tissues is not well studied. Only a handful of reports touch on this subject.\n\nFlugel’s group tracked antigen-specific CD4 effector T cells in the mouse brain under conditions of experimental autoimmune encephalitis (EA)30. Although the antigen presenting cells (APCs) were not directly visualized, 35% of the T cells stopped and appeared to make an immunologic synapse suggested by polarized expression of LFA-1 and TCR. This was antigen-specific, because ovalbumin-specific T cells were only tethered after intrathecal injection of antigen. Carbone’s group studied memory T cell activation using reactivation by herpes simplex virus in dorsal root ganglia of mice31. Although no imaging was performed, the key finding is that resident T cells can expand locally in an antigen-specific manner. Bousso’s group studied Leishmania major antigen recognition in the skin and found that CD4 T cells can activate nearby cells by IFN-γ, which can diffuse as far as 80 μm32.\n\nSteinman’s group isolated vascular dendritic cells from CD11c-YFP mouse aortas and aortic valves and incubated them with transgenic T cells33. They found that these CD11c+ DCs expressed MHC-II, CD80 and CD86, but not CD40. They were able to cross-present protein antigens on MHC-I to CD8 T cells. In a second study, the same group showed that these vascular DCs are dependent on the growth factor receptor Flt3, have tolerogenic properties and some of them express the integrin CD10334. In a study of antigen presentation in the atherosclerotic aorta, my lab showed that antigen-experienced (CD44hiCD62L-), but not naïve CD4 T cells interact productively with APCs in the adventitia and the atherosclerotic plaque35. The APCs were visualized by green fluorescent protein (GFP) driven by the CD11c promoter36 and the T cells were labeled ex vivo to directly study interactions in the aorta. Three lines of evidence showed that the interactions were productive. First, long interactions between CD4 T cells and APCs were only observed when antigen was present. Second, the migration velocity of the T cells was drastically reduced during these interactions, from about 10 to about 3 μm/min. Third, cytokine production (IFN-γ, TNF) was observed when explanted aortas were incubated with antigen-experienced, but not naïve T cells. Taken together, these data show that a productive recall response is supported in the non-lymphoid tissue35. We did not investigate which co-stimulatory molecules were involved in these interactions of CD4 T cells with APCs. We also did not identify the precise nature and phenotype of the DCs. Also, we did not retrieve the T cells from the aortas to study their commitment to Th1 or Th17 or other subsets. Finally, we did not study chromatin remodeling in these T cells.\n\nThe second touch hypothesis, if correct, allows several testable predictions:\n\nIncompletely committed T cells should be found in lymph nodes and/or in efferent lymph. These cells would express homing receptors like chemokine receptors and adhesion molecules, but would not be fully committed to T1, Th2, Th17 or Treg (i.e., lack epigenetic markers).\n\nT cell polarization would be expected to be more complete after interaction with APCs in non-lymphoid tissues. This could be tested by T-bet expression for Th1, RORγt expression for Th17 or FoxP3 expression for iTreg, and by looking for the epigenetic signatures of full commitment.\n\nA complete set of polarizing signals may be especially critical for iTregs, because they need to come in contact with active TGF-β produced by the same cell that presents the relevant antigen, and active TGF-β requires αVβ8 integrin expression on that same cell, which may be an M2 tissue-resident macrophage.\n\nBlocking CD28 or CD80 and CD86 in non-lymphoid tissues would be expected to only have a limited impact, because TNF superfamily molecules are expressed on APCs in non-lymphoid tissues and can serve as alternative co-stimulatory molecules.\n\nAn interesting example of CD4 T cell interaction with APCs in a non-lymphoid tissue is “licensing” of encephalitogenic T cells in the lung as described in37. The details of this “licensing” process are not known. The licensing process is not antigen-specific.\n\n\nChromatin remodeling and phenotypic stability\n\nCD4 T-helper cells can interconvert between phenotypes. In general, iTregs and Th17 cells are thought to be less stable and Th1 and Th2 cells are thought to be more stable. A more stable phenotype is associated with chromatin remodeling in the region of key cytokine and transcription factor genes. This is best described for Th1 and Th2 cells.\n\nThe histone 3 lysine 4 trimethylation (H3K4me3) mark is associated with a permissive chromatin state, whereas H3K27me3 is suppressive. The H3K27me3 repressive tag is effectively removed by the enzyme encoded by Jmjd3, which is induced by T-bet, the defining transcription factor of Th1 cells. This includes removal of H3K27me3 from the Ifng locus38,39. To some extent, T-bet also represses genes that would be expressed in alternative T-helper fates, like Bcl6, the defining transcription factor of TFH cells15. However, FOXP3 and T-bet can be co-expressed in the same CD4 T cell40. Similarly, some FOXP3+ T cells secrete IL-17A41. There is no accepted nomenclature for these “Th1 Tregs” and Th17 Tregs” yet.\n\nThe genes encoding the Th2 cytokines IL-4, IL-5 and IL-13 are found on mouse chromosome 11 very close to each other. This locus is extensively remodeled in Th2 cells. A 3’ enhancer of IL-4 expression called HSV is required for stable activation of the IL-4 promoter42, especially in TFH cells. The epigenomic modifications leading to iTreg43, Th17 and TFH are under investigation.\n\nIn the context of the second touch hypothesis, it would seem reasonable to postulate that the second touch is required for epigenomic modifications and stable T-helper phenotype commitment.\n\n\nEvidence supporting the second touch hypothesis\n\nThe finding that induction of most homing receptors are T-bet dependent15 supports the hypothesis that acquisition of homing receptors may precede full T cell differentiation: T-bet is the defining transcription factor of the Th1 lineage, yet the homing T cells can be of any lineage. As a specific example, the Th1 chemokine receptor CXCR3 is induced rapidly, and the peak of its expression precedes maximal T cell expansion14.\n\nThe expression profile of co-stimulating receptors (TNF superfamily members TNFSF and TNF receptor superfamily members TNFRSF) and cytokines (IL-12, IL-23, TGF-β) is different in macrophages and DCs in inflamed tissues compared to DCs in secondary lymphoid tissues44. This finding supports the idea that the signals received by antigen-experienced CD4 T cells cruising through inflamed tissues are different from those received in secondary lymphoid organs. Myeloid cells in tissues are mostly macrophages, which are usually M2 polarized, secrete active TGF-β and thus promote Treg induction. In inflamed tissues, most macrophages are M1-polarized, secrete IL-12 and IL-23 and thus support Th1 and Th17 cell development.\n\nOnly a very small fraction of DCs are migratory. These few cells are unlikely to fully represent the entire cytokine environment of the inflamed tissue. Certainly, resident tissue macrophages, a cell type known to be M2 polarized and anti-inflammatory, are not migratory45. It follows that it is likely that antigen-experienced T cells would encounter a different cytokine milieu and different co-stimulating receptors in the inflamed tissue itself rather than in the secondary lymphoid organs.\n\nTissue macrophages are M2-polarized and produce TGF-β. However, TGF-β is secreted in an inactive form and requires αVβ8 integrin for activation27–29. The activating integrin must be expressed on the same cell that presents the antigen, suggesting that only TGF-β secreted by αVβ8 integrin-expressing macrophages and DCs is relevant for iTreg induction. Therefore, only M2 resident macrophages (which cannot migrate to secondary lymphoid organs) in tissues would be expected to be effective at producing iTregs. There is good evidence that iTregs from non-lymphoid tissues are the ones that are relevant in curbing excessive inflammation10. Therefore, one of the most striking consequences of the second touch hypothesis would be that only iTregs that circulate back into the tissue are effective at curbing inflammation, because it is only in the context of M2 tissue macrophages that they become effectively committed to the iTreg lineage.\n\nIn inflamed tissues, pro-inflammatory M1 macrophages, both converted from resting tissue macrophages and derived from newly recruited monocytes, express their own characteristic cytokines including IL-12 and IL-23. The vast majority of M1 macrophages remain in the inflamed tissue, but some leave and enter the draining lymph node. Under inflammatory conditions, the difference between dendritic cells and M1 macrophages blurs, because both are monocyte-derived46,47.\n\n\nConsequences of the second touch hypothesis\n\nIf the second touch hypothesis is correct, it follows that activated antigen-experienced CD4 T cells are “immature” effector T cells when they first leave the priming lymph node. They are immature, because they are not fully committed to a differentiation program yet. To fully commit, they would require the second touch of seeing antigen presented by APCs in the inflamed tissue, providing different co-stimulators, perhaps in the form of interactions between TNF superfamily members with their receptors, or via soluble cytokines, or both. The “immature” effector T cell may be on the path to differentiation, but may not have undergone chromatin remodeling, which is needed for stable T helper phenotypes.\n\nIn addition, the nature and location of the second touch should also influence memory T cell polarization. This is well documented in two cases. Memory cells derived from TFH cells have the phenotype of central memory (TCM) cells48, because they reside in the secondary lymphoid organs and home back there (through CCR7, CD62L and other homing mechanisms). Conversely, effector T cells recirculating through an inflamed tissue (via CXCR3, CCR4, CCR9, α4β7 integrin or PSGL-1) would be expected to have a phenotype of effector memory T cells (TEM). They may show Th1, Th2 or perhaps Th17 polarization. Specifically, T cells that encountered antigen in the intestinal tract acquire expression of the α4β7 integrin and preferentially home back to the intestine12,49.\n\n\nCo-stimulation by TNF and TNF receptor superfamily members\n\nCo-stimulation occurs not only through C28 binding CD80 and CD86, but also through TNF superfamily molecules on T cells binding TNF receptor superfamily molecules on APCs50 (Figure 3). The main co-stimulator molecules of this class are listed in Table 1.\n\nThe first touch occurs in the lymph node, where the naïve T cell (light blue) is exposed to antigen (green circle) presented by MHC-II (purple) in the context of CD80 and CD86 (orange) co-stimulation, which bind CD28 (pink). The MHC-II-peptide complex is recognized by the T cell receptor (TCR, yellow) with CD4 (green). The second touch occurs in the tissue, where antigen is presented to antigen-experienced (orange) T cells by M1 (pink) and M2 (green) macrophages. Co-stimulatory molecules are likely from the TNFSF (red) on the T cell and the TNFRSF (blue) on the APC (see table for list of molecules). Characteristic M1 cytokines like IL-12 and IL-23 commit the CD4 T cell to Th1 and Th17, respectively. The M2 cytokine TGF-β commits T cells to the iTreg lineage.\n\nAll five TNFSF family co-stimulatory molecules are constitutively expressed on Tregs. These and many of the TNFSFR members are induced on other T cells at 24 hours after antigen presentation. Engagement of these receptors increases IL-4 and IFN- production. OX40 promotes effector T cells and blocks the development of iTregs. Inhibiting any of the four listed pairs ameliorates many models of autoimmune disease. Taken together, the TNFSF and TNFRSF molecules are possible candidates for the second touch signal that T cells may receive in the inflamed tissue.\n\n\nConclusion\n\nIn conclusion, the second touch hypothesis accommodates the plasticity of CD4 T cell phenotypes. It fills in the disconnect between the location of naïve T cell activation (PLN) and T cell effector functions (inflamed tissues). The concept of the second touch hypothesis generates many sub-hypotheses that can be tested by manipulating the mouse genome in a cell-specific fashion. Finally, the second touch hypothesis generates a conceptual framework for successful translation of immunological concepts to medical applications. This would include protective and tolerogenic vaccines and biologics that manipulate cytokines and their receptors.", "appendix": "Competing interests\n\n\n\nThe author declares 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\nGermain RN, Miller MJ, Dustin ML, et al.: Dynamic imaging of the immune system: progress, pitfalls and promise. Nat Rev Immunol. 2006; 6(7): 497–507. PubMed Abstract \| Publisher Full Text\n\nMempel TR, Henrickson SE, von Andrian UH, et al.: T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature. 2004; 427(6970): 154–159. PubMed Abstract \| Publisher Full Text\n\nCatron DM, Itano AA, Pape KA, et al.: Visualizing the first 50 hr of the primary immune response to a soluble antigen. Immunity. 2004; 21(3): 341–347. PubMed Abstract \| Publisher Full Text\n\nTacke F, Ginhoux F, Jakubzick C, et al.: Immature monocytes acquire antigens from other cells in the bone marrow and present them to T cells after maturing in the periphery. J Exp Med. 2006; 203(3): 583–597. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMills CD, Kincaid K, Alt JM, et al.: M-1/M-2 macrophages and the Th1/Th2 paradigm. J Immunol. 2000; 164(12): 6166–6173. PubMed Abstract\n\nItano AA, Jenkins MK: Antigen presentation to naive CD4 T cells in the lymph node. Nat Immunol. 2003; 4(8): 733–739. PubMed Abstract \| Publisher Full Text\n\nRen Y, Stuart L, Lindberg FP, et al.: Nonphlogistic clearance of late apoptotic neutrophils by macrophages: efficient phagocytosis independent of beta 2 integrins. J Immunol. 2001; 166(7): 4743–4750. PubMed Abstract\n\nLucas M, Stuart LM, Savill J, et al.: Apoptotic cells and innate immune stimuli combine to regulate macrophage cytokine secretion. J Immunol. 2003; 171(5): 2610–2615. PubMed Abstract\n\nChistiakov DA, Sobenin IA, Orekhov AN: Regulatory T cells in atherosclerosis and strategies to induce the endogenous atheroprotective immune response. Immunol Lett. 2013; 151(1–2): 10–22. PubMed Abstract \| Publisher Full Text\n\nJosefowicz SZ, Lu LF, Rudensky AY: Regulatory T cells: mechanisms of differentiation and function. Annu Rev Immunol. 2012; 30: 531–64. PubMed Abstract \| Publisher Full Text\n\nShiow LR, Rosen DB, Brdickova N, et al.: CD69 acts downstream of interferon-alpha/beta to inhibit S1P1 and lymphocyte egress from lymphoid organs. Nature. 2006; 440(7083): 540–544. PubMed Abstract \| Publisher Full Text\n\nIwata M, Hirakiyama A, Eshima Y, et al.: Retinoic acid imprints gut-homing specificity on T cells. Immunity. 2004; 21(4): 527–538. PubMed Abstract \| Publisher Full Text\n\nLey K, Kansas GS: Selectins in T-cell recruitment to non-lymphoid tissues and sites of inflammation. Nat Rev Immunol. 2004; 4(5): 325–335. PubMed Abstract \| Publisher Full Text\n\nGroom JR, Richmond J, Murooka TT, et al.: CXCR3 chemokine receptor-ligand interactions in the lymph node optimize CD4+ T helper 1 cell differentiation. Immunity. 2012; 37(6): 1091–1103. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOestreich KJ, Weinmann AS: Transcriptional mechanisms that regulate T helper 1 cell differentiation. Curr Opin Immunol. 2012; 24(2): 191–195. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJenkins MK, Chu HH, McLachlan JB, et al.: On the composition of the preimmune repertoire of T cells specific for Peptide-major histocompatibility complex ligands. Annu Rev Immunol. 2010; 28: 275–94. PubMed Abstract \| Publisher Full Text\n\nGermain RN, Jenkins MK: In vivo antigen presentation. Curr Opin Immunol. 2004; 16(1): 120–125. PubMed Abstract \| Publisher Full Text\n\nOh SA, Li MO: TGF-β: guardian of T cell function. J Immunol. 2013; 191(8): 3973–3979. PubMed Abstract \| Publisher Full Text\n\nShull MM, Ormsby I, Kier AB, et al.: Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature. 1992; 359(6397): 693–699. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarie JC, Liggitt D, Rudensky AY: Cellular mechanisms of fatal early-onset autoimmunity in mice with the T cell-specific targeting of transforming growth factor-beta receptor. Immunity. 2006; 25(3): 441–454. PubMed Abstract \| Publisher Full Text\n\nSakaguchi S, Yamaguchi T, Nomura T, et al.: Regulatory T cells and immune tolerance. Cell. 2008; 133(5): 775–787. PubMed Abstract \| Publisher Full Text\n\nZhou L, Chong MM, Littman DR: Plasticity of CD4+ T cell lineage differentiation. Immunity. 2009; 30(5): 646–655. PubMed Abstract \| Publisher Full Text\n\nZielinski CE, Mele F, Aschenbrenner D, et al.: Pathogen-induced human TH17 cells produce IFN-γ or IL-10 and are regulated by IL-1beta. Nature. 2012; 484(7395): 514–518. PubMed Abstract \| Publisher Full Text\n\nSallusto F, Lenig D, Forster R, et al.: Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999; 401(6754): 708–712. PubMed Abstract \| Publisher Full Text\n\nGebhardt T, Whitney PG, Zaid A, et al.: Different patterns of peripheral migration by memory CD4+ and CD8+ T cells. Nature. 2011; 477(7363): 216–219. PubMed Abstract \| Publisher Full Text\n\nHarrington LE, Janowski KM, Oliver JR, et al.: Memory CD4 T cells emerge from effector T-cell progenitors. Nature. 2008; 452(7185): 356–360. PubMed Abstract \| Publisher Full Text\n\nAcharya M, Mukhopadhyay S, Paidassi H, et al.: αv Integrin expression by DCs is required for Th17 cell differentiation and development of experimental autoimmune encephalomyelitis in mice. J Clin Invest. 2010; 120(12): 4445–4452. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMelton AC, Bailey-Bucktrout SL, Travis MA, et al.: Expression of αvβ8 integrin on dendritic cells regulates Th17 cell development and experimental autoimmune encephalomyelitis in mice. J Clin Invest. 2010; 120(12): 4436–4444. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTravis MA, Reizis B, Melton AC, et al.: Loss of integrin alpha(v)beta8 on dendritic cells causes autoimmunity and colitis in mice. Nature. 2007; 449(7160): 361–365. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKawakami N, Nagerl UV, Odoardi F, et al.: Live imaging of effector cell trafficking and autoantigen recognition within the unfolding autoimmune encephalomyelitis lesion. J Exp Med. 2005; 201(11): 1805–1814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWakim LM, Waithman J, van Rooijen N, et al.: Dendritic cell-induced memory T cell activation in nonlymphoid tissues. Science. 2008; 319(5860): 198–202. PubMed Abstract \| Publisher Full Text\n\nMuller AJ, Filipe-Santos O, Eberl G, et al.: CD4+ T Cells Rely on a Cytokine Gradient to Control Intracellular Pathogens beyond Sites of Antigen Presentation. Immunity. 2012; 37(1): 147–157. PubMed Abstract \| Publisher Full Text\n\nChoi JH, Do Y, Cheong C, et al.: Identification of antigen-presenting dendritic cells in mouse aorta and cardiac valves. J Exp Med. 2009; 206(3): 497–505. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoi JH, Cheong C, Dandamudi DB, et al.: Flt3 Signaling-Dependent Dendritic Cells Protect against Atherosclerosis. Immunity. 2011; 35(5): 819–831. PubMed Abstract \| Publisher Full Text\n\nKoltsova EK, Garcia Z, Chodaczek G, et al.: Dynamic T cell-APC interactions sustain chronic inflammation in atherosclerosis. J Clin Invest. 2012; 122(9): 3114–3126. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindquist RL, Shakhar G, Dudziak D, et al.: Visualizing dendritic cell networks in vivo. Nat Immunol. 2004; 5(12): 1243–1250. PubMed Abstract \| Publisher Full Text\n\nOdoardi F, Sie C, Streyl K, et al.: T cells become licensed in the lung to enter the central nervous system. Nature. 2012; 488(7413): 675–679. PubMed Abstract \| Publisher Full Text\n\nHatton RD, Harrington LE, Luther RJ, et al.: A distal conserved sequence element controls Ifng gene expression by T cells and NK cells. Immunity. 2006; 25(5): 717–729. PubMed Abstract \| Publisher Full Text\n\nLee K, Gudapati P, Dragovic S, et al.: Mammalian target of rapamycin protein complex 2 regulates differentiation of Th1 and Th2 cell subsets via distinct signaling pathways. Immunity. 2010; 32(6): 743–753. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoch MA, Tucker-Heard G, Perdue NR, et al.: The transcription factor T-bet controls regulatory T cell homeostasis and function during type 1 inflammation. Nat Immunol. 2009; 10(6): 595–602. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoo KS, Wang YH, Santori FR, et al.: Identification of IL-17–producing FOXP3+ regulatory T cells in humans. Proc Natl Acad Sci U S A. 2009; 106(12): 4793–4798. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVijayanand P, Seumois G, Simpson LJ, et al.: Interleukin-4 production by follicular helper T cells requires the conserved Il4 enhancer hypersensitivity site V. Immunity. 2012; 36(2): 175–187. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLal G, Bromberg JS: Epigenetic mechanisms of regulation of Foxp3 expression. Blood. 2009; 114(18): 3727–3735. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCroft M: Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat Rev Immunol. 2003; 3(8): 609–620. PubMed Abstract \| Publisher Full Text\n\nJenkins SJ, Ruckerl D, Cook PC, et al.: Local macrophage proliferation, rather than recruitment from the blood, is a signature of TH2 inflammation. Science. 2011; 332(6035): 1284–1288. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGeissmann F, Manz MG, Jung S, et al.: Development of monocytes, macrophages and dendritic cells. Science. 2010; 327(5966): 656–661. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSerbina NV, Salazar-Mather TP, Biron CA, et al.: TNF/iNOS-producing dendritic cells mediate innate immune defense against bacterial infection. Immunity. 2003; 19(1): 59–70. PubMed Abstract \| Publisher Full Text\n\nCrotty S: Follicular helper CD4 T cells (TFH). Annu Rev Immunol. 2011; 29: 621–63. 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\nCroft M: The role of TNF superfamily members in T-cell function and diseases. Nat Rev Immunol. 2009; 9(4): 271–285. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRescigno M: Functional specialization of antigen presenting cells in the gastrointestinal tract. Curr Opin Immunol. 2010; 22(1): 131–6. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4185", "date": "25 Mar 2014", "name": "David Voehringer", "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 opinion articles provides a hypothesis that T cells are only weakly activated in lymphoid organs but required a \"second touch\" in inflamed tissues to stabilize their differentiation program into different T effector cell subsets.The article is well written but would benefit from removing redundant information regarding the different T cell subsets and M1/M2 macrophages which are given at several places in the manuscript. Further, most published data indicate that M1/M2 macrophages are induced by Th1/Th2 cells rather than M1/M2 polarization being regulated by Th1/Th2 cells. The notion that basophils induce Th2 cells should be removed as recent reports do not support this finding.", "responses": [] }, { "id": "4184", "date": "25 Mar 2014", "name": "Lawrence Patrick Kane", "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 opinion piece by Ley is interesting and well-timed. Thus, the topic of peripheral T cell activation and differentiation is well-trodden ground at this point, but discussions of these events often stop at the border of the lymph node, where priming occurs. Here, Ley discusses the accumulating evidence that activated T cells require additional signals after their arrival in peripheral non-lymphoid tissues, in order to become fully differentiated. This is a topic that has received much less attention, and warrants further attention from T cell biologists. I have three major suggestions for revision of the piece as it stands now:The author dedicates approximately one-third of the article at the beginning to reviewing various aspects of T cell priming in lymph nodes. In my opinion, this runs the risk of losing the reader’s interest, before reaching the more interesting discussion of the “second touch” hypothesis. I would suggest significantly trimming this part of the article, since it has been so thoroughly reviewed elsewhere. In addition, the author could circle back to this after the more novel discussions of peripheral differentiation. It would also be helpful to hear more about previous studies on TNF/TNFr family members and T cell differentiation, in the last section of the article (there should be more space if the first part is shortened, as suggested above). Finally, I found the Table listing TNF/TNFr family members somewhat confusing. Does the heading “T cell” at the top refer to “constitutive expression on T cells”? The headings should be clarified or perhaps additional columns could be used for listing expression patterns. Other, more minor, comments:Page 6, second column – CD28 is misspelled as “C28”.A few sentences could use additional editing to make them easier to read (e.g. page 6, first column has two such sentences).", "responses": [] }, { "id": "4188", "date": "03 Apr 2014", "name": "Luc Teyton", "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 make two major comments:The title will be confusing for a lot of readers as it is reminiscent of the two signal hypothesis, which concerns naive T cell activation and the requirement for a co-stimulatory signal.Here, the issue is very different since the whole idea is that what happens to a T cell in the lymph node and in the inflamed tissue is substantially different. Here, everyone will agree, the set of effector functions that is displayed in the lymph node is meant to produce differentiation and proliferation. Whereas in inflamed tissue the effector functions will be the destruction of infected cells.In other words, the question could instead be asked as \"Why do we have lymph nodes?”. Here, the discussion and hypothesis could go wild - but the concept of concentrating immune cells in a constrained space to optimize communications is very attractive. Tissue versus lymph node - I agree with the author when he says that not enough is known about the trafficking of T cells to lymph nodes and tissues. Which comes first?It is counter-intuitive to follow the idea that priming of T cells is exclusively a lymph node process. Indeed, the molecular cues that allow T cells to enter tissues are coming from the tissue itself and the site of inflammation, not the lymph node. It is much easier to understand the story of the monocyte/dendritic cell if we consider the resident population - it will be at the site of injury and will receive the information to pickup antigen and migrate to lymph nodes.One thing that the author may want to consider in a revised version is the role of tissue anatomy in the circulation and recirculation of lymphocytes in tissues. Much of the ingress of lymphocytes into tissues has been studied from the capillary side. However, in tissues, lymphocytes are concentrated in the lymphatics, not in the capillary end of blood circulation. When taken into consideration, this factor appears essential for the priming in lymph nodes. In other words, what requires further study is the concentration of lymphocytes in tissue lymphatics and the demonstration that molecules such as TNF family members participate in the first phase of priming. Minor comments: the description of the various T cell subsets (twice) is distracting from the important question - it should be simplified or simply referenced.", "responses": [] }, { "id": "4253", "date": "14 Apr 2014", "name": "Diana Boraschi", "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 my opinion, this paper is important for two reasons. First, it builds on available data, looking in particular at discrepancies with the current theories, for building a new hypothesis that reconciles most of the unexplained results obtained in the past and were somehow forgotten or set aside. Secondly, by doing so it provides a very good example of how scientific thinking should work, a lesson that many young (and also less young) researchers need to learn. Research is going against dogmas, when dogmas do not explain real life, and formulating new hypotheses, which will the drivers of future research. In this particular case, the hypothesis that T cell polarisation/differentiation does not occur all at once, but that the terminal direction is reached in the specific site of the ongoing defense reaction is indeed a logical assumption, thus likely to be true. An important aspect of this paper is that the author puts the immune reaction right into the geographical and temporal framework of its development. Thus, where and when an immune response takes place could make a significant difference in its outcomes.The paper is well written. The initial summary of current knowledge (which another reviewer, Lawrence Kane, has found too long) is indeed long, but not useless. It may be shortened but it is good to provide a clear picture of what is the current knowledge and hypothesis. Also, in full agreement with Lawrence’s comment, the author should emphasize the description of the second touch hypothesis, which in this version comes late and almost incidentally. I also agree on finding Table 1 confusing (I couldn’t even understand its title).I am particularly pleased, as my major interest is in macrophages, that this hypothesis resurrects the role of macrophages as antigen-presenting cells, a role that has been for a long time forgotten with Steinman’s “marketing” of DC as the one and only APC. Macrophage polarisation into M1 and M2 is briefly touched in this paper, but this is OK since this was not a central issue. However, it may be worth remembering that M2 polarisation comprises, in the current terminology, all that is not M1, and therefore there are the inflammatory M2 macrophages that inform/support Th2 responses (alternative/type 2 inflammatory cells); the anti-inflammatory, tissue repair promoting cells (those that dump all types of inflammatory reactions, both type 1 and type 2, and produce matrix component, angiogenic and proliferation-inducing factors, etc.); and a heterogeneous group of intermediate functional phenotypes. Where exactly tissue resident macrophages fall in this description is still not clarified but since their reaction to any change in the tissue microenvironment is practically immediate, they will be functionally different depending on the tissue conditions - inflammation being a violent and massive kind of change in the tissue microenvironment.I partly disagree with David Voehringer's comment that M1/M2 polarisation is dictated by Th1/Th2 cells, while Klaus in his paper state the reverse. Charles Mills et al. (2000) demonstrated how M1/M2 macrophages can actually direct T cell polarisation, opposite to the common belief that Th1/Th2 do regulate macrophage polarisation. I guess that the truth may be half way (in medio stat virtus), since there are clear data that macrophages induce T cell polarisation and also data showing that macrophage polarisation is induced/amplified by polarised T cell cytokines. In a tissue during an ongoing inflammatory reaction, both things would have the possibility of occurring. We should also be aware that monocytes and lymphocytes infiltrating the site of inflammation do not come all at once at a single time point, but instead continuously during the entire course of the reaction - thus the antigen-experienced cells and the blood monocytes find, depending on when they arrive, completely different microenvironments and could therefore undergo different differentiation/regulation.To conclude, I really liked the paper, not only its content but also its ability to stimulate my interest and wish to know more and go deeper.", "responses": [] } ]	1	https://f1000research.com/articles/3-37
https://f1000research.com/articles/3-71/v1	13 Mar 14	{ "type": "Research Article", "title": "Developing sustainable software solutions for bioinformatics by the “Butterfly” paradigm", "authors": [ "Zeeshan Ahmed", "Saman Zeeshan", "Thomas Dandekar", "Saman Zeeshan" ], "abstract": "Software design and sustainable software engineering are essential for the long-term development of bioinformatics software. Typical challenges in an academic environment are short-term contracts, island solutions, pragmatic approaches and loose documentation. Upcoming new challenges are big data, complex data sets, software compatibility and rapid changes in data representation. Our approach to cope with these challenges consists of iterative intertwined cycles of development (“Butterfly” paradigm) for key steps in scientific software engineering. User feedback is valued as well as software planning in a sustainable and interoperable way. Tool usage should be easy and intuitive. A middleware supports a user-friendly Graphical User Interface (GUI) as well as a database/tool development independently. We validated the approach of our own software development and compared the different design paradigms in various software solutions.", "keywords": [ "Typical challenges in bioinformatics in an academic environment include “ad hoc” programming. No maintenance is really possible as scientists such as PhD students and post-doctoral scholars leave after their thesis is completed or after their post-doc contract. These scientists may also have no formal computer science training", "and often there is no structured programming1 and solutions might not be compatible with each other. Furthermore", "in an academic environment there are a number of inherent pressures to develop pragmatic and fast (“quick and dirty”) software solutions2." ], "content": "Introduction\n\nTypical challenges in bioinformatics in an academic environment include “ad hoc” programming. No maintenance is really possible as scientists such as PhD students and post-doctoral scholars leave after their thesis is completed or after their post-doc contract. These scientists may also have no formal computer science training, and often there is no structured programming1 and solutions might not be compatible with each other. Furthermore, in an academic environment there are a number of inherent pressures to develop pragmatic and fast (“quick and dirty”) software solutions2.\n\nIn addition, there are some new and “modern” challenges, which become more and more pressing simply as the technology progresses: big data3, the wave of “omics” data to process, and the problem of interoperability of software tools4. Well-known recent solutions for this challenge are Taverna5 and Galaxy6,7. The latter in particular is well suited to dealing with large quantities of data such as new large-scale sequencing data.\n\nAnother issue is that the data should be accessible, with uniform syntax and rich semantics for integration. Furthermore, data schemes are prone for change due to rapid advances in the field, so a schema-free representation of data is increasingly important (for scientific data). The UniProt Consortium8 for instance has recently shifted from the use of relational databases to the semantic web for flexible data management.\n\nTo counter these older and general as well as new challenges, we have now developed a solution of iterative and intertwined development cycles (Butterfly model), which improves the typical aspects of long-term sustainability and maintenance. Furthermore, it features detailed user-requirement analysis, good graphical and simple user interfaces (optimized human-computer interface, HCI) and intuitive software use that also exploits natural language processing. Our model tackles current challenges: first, the interoperability of the software takes into account middleware solutions, so that both database and user interfaces can be used flexibly. Second, the Butterfly approach improves the meticulous database engineering required to build large-scale and “omics” databases.\n\nThe Butterfly development cycle is different from previous approaches9–12. It requires some additional time investment at the start for design and implementation, but this pays off later. The Butterfly model worked well in our hands regarding the above challenges (section: Real time examples using Butterfly). In the following we confirm this by describing concrete software development.\n\nThe basic concept is really simple: plan ahead, back-check the critical development steps by a separate sub-cycle and talk with the user. It is important to spend more time on requirement analysis, as well as to invest well in interoperability and maintenance. We do not claim that these problems have not been recognized before nor that no alternative solutions for this are available2,6–8, but we are confident that with our approach it will be possible to obtain particularly well optimized and high quality bioinformatics software solutions in an academic environment. The initial time investment in the Butterfly paradigm helps to save time later due to the interoperability and easy maintenance of the software solutions achieved.\n\nThis paper is organized as follows: this section sets the stage for our agenda (section 1); Current Software Engineering and Development (section 2) highlights the modular phases of software development processes and compares several typical software strategies highlighting the novelty of our approach; next, the Butterfly work flow is explained (section 3), and software examples using the Butterfly approach (section 4) validate the Butterfly design principle by providing concrete examples of software projects from own work. Moreover we discuss some bioinformatics tools based on their type, methodology and usage.\n\n\nCurrent Software Engineering and Development\n\nSoftware Engineering (SE)9 is one of the most recognized fields in computer science as it matures and expedites the processes of software development. In particular, it allows a focus on the life cycle of software and sustainable development as an improvement to pragmatic short-lived implementations. SE has introduced many process improvement models and techniques10–12, and Software Development Life Cycle (SDLC) models13, with some variabilities14 and commonalties15–19. In general, depending upon the observed commonalties, we state\n\n“Software Engineering is an integrated, cyclic and product line combination of following independent modular approaches: requirements engineering20–22, design modelling23–25,27–35 programming, testing and deployment”.\n\nThe five modular SE approaches remain the same when it comes to the software engineering of the scientific software solution development (Figure 1). However, in contrast to a pragmatic and maybe traditional software application development in an academic setting (Figure 2), a major change is the inconsistency in all phases of the SDLC. In the requirement engineering phase (Figure 2; traditional software solution development), all requirements should be provided before the start of design. This is not the case when dealing with most of the scientific software applications, and the requirements continuously change with the passage of time (we have proposed an updated SSE SDLC Model, Figure 1; scientific software solution development). Ultimately, this complicates the process of analysis and filters out functionals. Programming structures become complex (Figure 1), as the possibilities of error proneness (both logical and syntax errors) increase due to the continuous increment of variabilities in the pre-processed source code15–19.\n\nSSE integrates and combines in a development cycle the following independent main modular approaches: requirements engineering, design modeling, programming, testing and deployment. Each approach consists of its own sub-modular, integrated and cyclic combination of internal phases: requirement engineering consists of specification, functionals, non-functionals, and analysis; design modeling consists of use cases, system flows, data flow and source code; programming consists of languages, tools and technologies, development, and debugging; testing consists of test cases, modular, integrated and quality; finally, deployment consists of installation, configuration, training, feedback. Iterative cycles lead to continuous improvement. Achievements translate the goals in good software.\n\nTesting of integrated and individual modules becomes time consuming (Figure 1), as new test cases have to be continuously rewritten and their application often leads to ‘ripple effects’29: these are unidentified logical or syntax errors in the system which arise while fixing the errors36,37. Depending upon the nature of the system, many approaches have been proposed to improve software quality control processes38–50 which improve standard software development and are important in scientific software quality assurance and improvement. Furthermore if the system keeps changing and is inconsistent, then the deployment procedures can also be complex and time consuming, especially for large applications with multiple interfaces and controls providing numerous individual and integrated functionalities.\n\nTo further help (SSE, non-computer scientist bioinformaticians etc.) in expediting the processes of adopting the concepts of SDLC, we provide a tabular comparison between different SDLCs, based on their commonalities and variabilities (Table 1). This comparison is based on following twenty four defined comparative SDLC (authors’ initiated) measures: Software Engineering Approach9, Initial, Developmental Plan, Software Requirements Engineering, In Depth Requirements Analysis51, Requirement Validation, Functionals, Risk Analysis, Software Design, Software Architecture Design, In Depth Software Design Modelling, Reusable Designing, Developmental Plan, Tools and Technology Selection and Analysis, Graphical User Interface Design, Preprocessed Source Code Writing15–18, Integrated Programming, Software Testing, In Depth Software Testing, Customer’s Evaluation, Deployment Procedures, Maintenance, Software Re-Engineering52, Cyclic or Repetitive, Easy to learn and Use, and User Training.\n\nWe applied these measures to nine different software development life cycle models: Waterfall Model53, V-Model54, Spiral Model55, Iterative and Incremental Model56, Rapid Prototype Model57, Agile Development Model58, Extreme Programming Model58, Evolutionary Model59, Code and Fix Model60. From this comparative analysis we conclude that there is no such one specific SDLC which can be helpful in all required phases of scientific software solution development, but some which might be more useful: Spiral, Waterfall and V-Model. Moreover the SDLCs famous for the quick development (Rapid Prototype Model, Agile Development Model, Extreme Programming Model, Evolutionary Model, Code and Fix Model), lack in quality software production.\n\nAs this is a general and open comparison, depending on the nature of the scientific software application, one can further analyze and pick that which suits best. Furthermore, we considered only the typical effort and strengths for each of these software development paradigms. A meticulous developer can of course take special care and spend more time on any of the features not typically covered by the software paradigm he follows, and turn the “no” for this feature into a “yes” simply by this additional effort during SDLC (for instance, regarding agile programming – for that matter, extreme programming can also be considered as a type of agile development). The goal for our “Butterfly” paradigm is a SDLC paradigm that fulfils all of the features regarding life cycle management of the resulting software.\n\n\nButterfly in scientific academia\n\nIf we search for “bioinformatics tools” over the web, thousands of entries can be found at one hit. But how many of those are fully designed, developed and tested solutions, used and maintained for a number of years and still in functional use?\n\nAt the beginning of a software project in academia, scientific solution development seems very interesting, fascinating and exciting. But with the passage of time, when the levels of complexities increase (due not only to the lack of developmental skills but also to the unavailability of proper designs), the work starts becoming tedious and unfruitful. This causes a lack of interest in software solution development and leads to a preference for wrapping up the work with a working script or small application, which can be published later on.\n\nHere we propose a new science-oriented model (Figure 3), which can help the scientific software solution developers as well as the scientists/end users by generalizing the use of major developmental aspects correlating to the important scientific needs of the target system. The name of our new model is “Butterfly”.\n\nIt consists of four wings: Scientific Software Engineering (upper left), Human Computer Interaction (lower left), Scientific Methodology (upper right) and Scientific Application (lower right). Moreover it leads to continuous improvement (in yellow). The achievements translate the goals into software.\n\nIn accordance with the name of the model, the ‘Butterfly’ represents sustainable and continuous improvements from goals to achievement as the central backbone of development. For larger projects this is further improved by using middleware between user interfaces and accessing the various data and databases involved. This also has the advantage that natural language processing needs to be implemented only once in the middleware and all the other modules have access to it. Finally, the backbone arrangement with a powerful middleware as well as all development elsewhere in the “Butterfly” stresses the interoperability of the software. It is developed by using well-defined and compatible output formats. Furthermore, below the middleware, well adapted, interoperable data schemes boost sustainable development of database structures, including efforts for scheme-less databases and other semantic web developments. The four continuously moving wings (Figure 3) represent the change in the requirements. The upper left wing of the model represents the scientific software engineering principles, and the lower left wing represents the HCI. The upper right wing represents the implementation of the specific methodology (focused on lowering the risk of development ripples) and the lower right wing represents the target scientific application producing the required results.\n\nSSE is the most important phase of the Butterfly model, which promotes the usage of any earlier mentioned SDLC involving requirements analysis, design modeling, programming and testing of the scientific software solution. It correlates with the phase Scientific Methodology; the finalized functional requirements are based on the desired system output, and the system should be modeled according to the defined logics and mathematics (individual as well as the sequence of algorithms if there are more than one). The most suitable, advanced, recent, economically affordable, transferable, flexible and reliable developmental technologies should be chosen considering the use and availability of the data (large, small, complex, shared via intranet or internet).\n\nMeanwhile, programming and processing of the complex and large data should be undertaken in order to have efficient data analysis, management and visualization. During the testing procedure of the developed system, all modules should be properly tested by the developer, by testing experts - if available -, and by the appropriate users. While testing the newly developed scientific system, the system should not be considered ready or functioning straight away. No real experiments should be performed prior to through testing to avoid any loss of data or waste of any scientific research/biological material or living beings (especially in case of behavioral research and analysis). Only after successful deployment the real time results should be evaluated.\n\nFor instance, if the target scientific software solution is a database manipulation and management system, then it will require to properly model the database schema (entity relationship model), by reducing the levels of data redundancy and dependency, via data normalization. There are five data normalization forms: 1NF, 2NF, 3NF, 4NF and 5NF, which include conceptual procedures for comprehensive database designing61. These data normalizations help in shaping the data types (1NF), developing the relationships between non-key and key fields (2NF, 3NF)62,63, and dealing with multi-valued facts which correspond to many relationships (4NF and 5NF)64,65.\n\nThe lower left wing of the Butterfly (Figure 3) is the HCI, known as Human Machine Interaction (HMI)66–68. This interacts with the lower right wing i.e. Scientific Application. HCI defines the implementation of the mechanisms that establish the efficient communication protocols between human and machines. These protocols are based on the textual, visual, sensory, video, audio and event based information, provided by both the user and the machine (computer). The backbone of the Butterfly allows by its middleware to rapidly exchange GUI applications and accessed databases if there is a need for it, considering the rapid developments in bioinformatics or if a user wishes to use GUIs or databases from comprehensive software environments, for instance regarding large-scale sequencing from Taverna5 or Galaxy6,7.\n\nUnfortunately HCI is the most ignored and unattended phase of scientific software solution development. Often developers do not give priorities to the GUI design and implementation. The reasons for this negligence could be the pressure due to time limits for development, rapid functionality addition during development, excessive iterations, less field knowledge, lack of awareness about the importance of HCI, competitive general purpose software and human behavior analysis etc.66. In principle and practice, with respect to the user’s point of view, HCI is one of the most important parts of the software development. If the HCI is bad and the software is not easy to use, why use the software at all? Below we will present a particularly well-engineered solution in this respect, the ant database, focusing on a really easy user database access by smart phone.\n\nDuring the scientific solution deployment and usage, the end user is probably a scientist without strong informatics background. To run and execute a script, first the compiler and the editor need to be installed, then depending upon the needed external libraries, the necessary libraries will need to be identified and used. This might be a very hectic task, especially for a person with no informatics background. In most of the cases, the result-producing applications cannot be used because of difficulties in deployment and execution procedures. Moreover, most of the time, many scientific software applications are not well documented, which also increases the level of complexities at the user end.\n\nIf the application is configured and is executed successfully, then the next task is to use it. If the application has been developed with an unfriendly GUI or a command line interface, it might be a problem for the user scientist, as to how to use it. If the application is not easily deployable and useable, then in most of the cases it can only lead to the loss of developer’s efforts.\n\nTo avoid such problems, the following aspects should be considered: user psychology, scientific domain, working environment and HCI design patterns. It is very important to understand the common psychology of the users of the applications e.g. if they are the laboratory scientists, they will be happy to have user friendly graphical interface and easy deployment procedure (e.g. small setup which runs and automatically configures all the required settings in the user’s system), so that they do not have to spend additional time on the configuration.\n\nScientific domain and working environments are two more important aspects to be considered while designing the graphical interface. These are particularly important especially when developing real time systems e.g. embedded, robotic, mobile applications. Moreover, it is important to consider that the use of HCI patterns and principles have a reliable HCI communication protocol implementation.\n\nWithout a doubt, designing a HCI interface meeting end user requirements is a very challenging and complex task. To reduce the level of complexities, HCI promotes the concept of a structured approach i.e. task decomposition. It helps developers make abstract designs with the use of specific processes and design artifacts, by keeping note of user requirements, usage scenarios and essential use cases. One of the best strategies to design an interface is first planning, then drawing sequential work-flow and finally writing down the levels of details.\n\nAfter designing an interface, one of the important tasks is to evaluate its effectiveness and potential. One general and effective way is to engage the users and consider their feedback at every step. The technical approach is to use the HCI design principles: Experimentation, Contextualization, Iteration and Empirical Measurement67. Another beneficial method is to adopt and use the HCI design patterns: Window Per Task, Direct Manipulation, Conversational Text, Selection, Form, Limited Selection Size, Ephemeral Feedback, Disabled Irrelevant Things, Supplementary Window and Step-by-Step Instructions67.\n\n\nButterfly workflow designs\n\nThe Butterfly model implementation mechanism has a three-layered architecture: Gray, Yellow and Green (Figure 4).\n\nShown in gray is the abstract layer, in yellow is the basis for design and development, and in green the implementation and testing by the user. The software is released including installation and training.\n\nThe gray layer represents the most important phase of scientific software solution development, which involves software designers, developers, testers, graphical interface designers and, most importantly, the users. It consists of four phases: scientific software solution planning, requirements engineering and analysis, conceptual software design and modeling, and user friendly graphical interface. The layer has been named ‘Gray’ because at the beginning of a new scientific software solution development, most of the information seems uncertain.\n\nScientific software solution planning (Figure 5) is the first step towards a new scientific application development, which requires the introduction to the field itself (e.g. biochemistry, neurobiology, genetics, metabolomics, proteomics etc.) and project related information (e.g. what could be the end product, input to the system, expected output from the system, methodology, ideas, opinions etc.). It is important to know about the user’s information IT background and existing already available (old and recently developed) scientific solutions to the problem. The next important phase is to perform requirements engineering and analysis (Figure 6). During this phase, the most important tasks are to gather the requirements from users (e.g. interviews, brainstorming, documents, publications, running related systems based information etc.) and classify the information in functional and non-functional categories. Here, function requirements are those which can be implemented (based on existing resources, time, budget, labor, tools, technologies and methodologies), and non-functional requirements are those which cannot be implemented. It is very important to clarify with the users what will be the expected end-product because it is possible that the user may not like the output of the system after development. In this case, all the efforts will have been in vain.\n\nAbstract planning is the first step of the top, abstract layer (gray, Figure 4) of the Butterfly design, key steps are indicated.\n\nThis is the 2nd step of the top layer (gray layer) of the three layer model in the Butterfly design. Key tasks are indicated.\n\nThe third phase is the conceptual software design and modeling (Figure 7). This is particularly important when there is a team of software developers. Before moving ahead, one should go for some abstract designs based on functional requirements and discuss these with the team. It is crucial to estimate the expected workflow, data sources and data flow in the system. If possible, the abstract design should be discussed with the users as well.\n\nThis is the 3rd step of the top layer in the butterfly model.\n\nThe last phase of the gray layer concerns the design of a user-friendly GUI (Figure 8). First, some mock-ups should be made (hand-made on papers or better to use white board with color markers and later make pictures of finalized GUI designs), then these should be discussed with users in a brain storming session. Finally, based on the perceived designs, the abstract GUI (GUI with no running functionalities) should be created using HCI design patterns.\n\nThis is the final step of the top, abstract layer (gray) of the three layers in the butterfly design.\n\nThe yellow layer involves designers and developers. It consists of four phases: design, modeling and analysis, tools and technology selection, design implementation, and GUI implementation.\n\nDuring design modeling and analysis, the important task is to create different implantable designs e.g. UML models (use case, class, system sequence, activity, component etc.) and database schemas (in case it’s a database management system). Here, we strongly recommend the use of Product Line architecture design modeling, where the whole software is divided into sets of modules, which work individually as well as together. This will customize project development and reduce error proneness during development. Moreover it will increase the concepts of modular reusability.\n\nThe next step involves the choice of available tools, technologies and programming languages that will be implemented in the designed models. The last step focuses on adding functionalities to the designed GUI.\n\nThe green layer describes the final in house testing and debugging by the developers and tester. Some scientific software applications are developed to process the raw data using mathematical algorithms (e.g. processing GC-MS, LC-MS, NMR data), whereas some applications are implemented to perform different kind of experiments, which in return produce experimental data e.g. towards behavioural research on animals and insects etc.\n\nProcessing raw data is safe even if there are still some problems (e.g. minor calculation mistakes due to the different levels of fractional values or wrong implementation of mathematical algorithms, or some software developmental problems issues, which could be the ‘ripple effects’, or some logical bugs) after testing. However, when applied in real time experimentation, if the software does not work as expected, it could be very expensive and dangerous (e.g. if there is a software to control the temperature and light during experiments on insects or animals, and there are problems during experimentation leading to changes of the normal or expected temperature and light to some extreme positive or negative values, then this could not only effect the system’s hardware but can also threaten the life of animals or insects).\n\nIn order to avoid such problems, we enforce test trails by different users before making the software available for installation and training in the public domains.\n\n\nReal time examples using Butterfly\n\nA number of new scientific software applications based on the concepts of the Butterfly model have already been proposed, designed, implemented, tested and are currently in use. Some of them have been published and some remain unpublished. These applications are: Least Square MIDA (LS-MIDA), DroLIGHT, Isotopo, Lipid-Pro and App Ant Database.\n\nLS-MIDA69,70 (Figure 9) is an own published scientific software (Department of Bioinformatics, Biocenter, University of Wuerzburg Germany) which estimates mass isotopomer distribution from the spectral data by analyzing each peak of given mass and each mass atom fragment. It implements a chain of mathematical and statistical algorithms, provides graphical interfaces to help users analyze experimental data and visualize the results, and provides a third party independent experimental data management system. For the most parts the requirement engineering and software design steps are followed until the final solution presented (Figure 9), and details on tested earlier solutions or specific requirements are available from the authors.\n\nScientific software solution towards bioinformatics and biochemistry which estimates mass isotopomers distribution from spectral data by analyzing each peak of given mass and each mass atom fragment. (http://www.tr34.uni-wuerzburg.de/computations/ls_mida/).\n\nDroLIGHT71–73 (Figure 10) is a published scientific computational solution towards neurobiology and photobiology (Department of Neurobiology and Genetics, Biocenter, University of Wuerzburg, Germany). It is a domain specific, intelligent, distributed, real time, embedded, data management system capable of controlling hardware devices, proficient in producing different colors of lights and monitoring the movements of Drosophila melanogaster. Moreover it also helps users generate circadian rhythms, provides a third party independent experimental data management system and produces experimentation output in two and three dimensional graphics formats. Constant improvement is further boosted by sustainable data structures and engineered robustness of the application. Furthermore, sustainable development and interoperability are considered from the start, support the rapid further development/improvement of software to study and manipulate circadian rhythm by different light cycles and test environments in drosophila.\n\nScientific software solution towards neurobiology and photobiology, capable of controlling and automating the hardware that produces different colors of lights via Light Emitting Diodes (LEDs). It provides experimental data management system, circadian rhythm generation and 3D visualization of system’s performance and experimentation details. (http://www.neurogenetics.biozentrum.uni-wuerzburg.de/en/project/services/drolight/).\n\nIsotopo software74 (Figure 11) is a published scientific software (Department of Bioinformatics, Biocenter, University of Wuerzburg, Germany), a bioinformatics solution with the ability of performing quantitative mass spectrometry in isotope labeling experiments. It is an extended version of the earlier software LS-MIDA with a well-optimized mathematics implementation. It not only provides the graphical interfaces for gas chromatography-mass spectrometry (GC-MS) experimental data analysis, visualization and management, but also provides an intelligent data parser to automatically transform the machine’s pre-processed data into the processable format of the Isotopo software (reducing both time and labor). It also provides a complete database management system as a simple, well sustainable version of a middleware between data storage and GUIs.\n\nScientific software solution towards bioinformatics and biochemistry, with the ability of performing quantitative mass spectrometry to mixtures of materials labeled with stable isotopes. It provides internal database management system, third party independent file based experimental data management system and intelligent data format parser for data extraction and conversion of different data formats. (http://spp1316.uni-wuerzburg.de/bioinformatics/isotopo/).\n\nLipid-Pro (Figure 12) is an unpublished scientific software (paper is in writing), a computational solution towards lipids and pharmaceutical biology, estimating fragment candidates, pre-parent candidates and most importantly lipids (Department of Pharmaceutical Biology, Biocenter, University of Wuerzburg, Germany). It implements a chain of mathematical operations and experimental data extraction processes. Along with the efficient third party independent data management system, it provides different graphical interfaces for individual and integrated intelligent operations.\n\nScientific software solution towards lipids and pharmaceutical biology, which estimates fragments, pre-parent candidates, fragment candidates and most importantly lipids. (http://www.neurogenetics.biozentrum.uni-wuerzburg.de/en/project/services/lipidpro/).\n\nApp Ant Database (Figure 13) is an unpublished scientific software (paper is currently in writing), featuring a distributed and embedded database system in the form of a smart phone, tablet and desktop application towards the rotating channel experiment data management during experimentation on desert ants (Department of Behavioral Physiology and Sociobiology, Biocenter, University of Wuerzburg, Germany). It is unique and the first bioinformatics smart phone application to be used in desert ant rotating channel experiments. After extensive requirement engineering, we established an extremely easy to use graphical interface. Furthermore, after studying the user requirements in monitoring desert ant movement and orientation in the desert, the application not only automatically records ant movements, but also estimates and calculates automatically all additional variables required for the project such as azimuth, solar time, equation of the time, time offset, hour angle, altitude, sunrise, sunset and solar noon using astronomical algorithms, recommended by the National Oceanic and Atmospheric Administration (NOAA).\n\nScientific software solution towards the rotating channel experiment data management during experimentation on desert ants. It offers user friendly graphical interfaces for the experimental data entrance, manipulation, management and sharing. (http://www.neurogenetics.biozentrum.uni-wuerzburg.de/en/project/services/ant_app_db/).\n\nRegardless of the individual specifications, development details, technologies used and usage perspectives, the key principles of the Butterfly model were applied in the design of these software solutions. Although the requirements for each of these were not fixed in the beginning, comprehensive requirements gathering and analysis operations were performed using brain storming and interviewing methods. Based on filtered out the functional requirements, the most suitable SDLCs (V-Model54 and Spiral55) were applied and the software applications were designed using UML (including use cases, class diagrams, sequence diagrams, work flows, activity flows, components diagrams, data flow diagrams etc.). Of those SDLC’s used in scientific software solution development, the Spiral Model proved most helpful and best suited due to its four main pillars: determine objectives, identify and resolve risk, development and test, and plan the next iteration. Another advantage of the spiral model is its risk driven approach, incorporating many useful features and refinements of other software development life cycle models54,72.\n\nUsing the HCI design patterns and principles, the graphical user interfaces of all these applications were designed considering the psychology, scientific and informatics backgrounds of the end users and the deployment environments.\n\nAll these applications are easy to deploy and use. We found that users did not require training to install, run and use the applications. As scientific research is a never ending process, these applications are still in development and will be continuously improved with respect to the methods, features, performance and technologies.\n\n\nComparing bioinformatics tools\n\nWe have performed a short comparative analysis of some bioinformatics software applications (C1375, Metatool76, BioOpt77, FiatFlux78 ReMatch79, Biolayout80, LS-MIDA69,70, DroLIGHT71–73, Isotopo74), describing their type, methodology, implementation, user friendliness, configuration etc., based on the provided, published information (Table 2).\n\nSSE=Scientific Software Application; App.=Application; DB=Database; DM=Data Management; Sys.=System; SDLC=Software Development Life Cycle.\n\nFor our comparison, we chose software applications from fields we are familiar with, yet tried to cover a broad range of different applications and compared Metabolic Flux Analysis (MFA) as well as software metabolic network analysis81,82, Mass Isotopomers Distribution Analysis (MIDA) including GC-MS data analysis69,70, and neurobiology applications for behavioral analysis of insects such as desert ants and fly71–73.\n\nWe used the following parameters to classify the chosen bioinformatics software application: SSE type, data management, script or prototype, algorithm type, algorithm/methodology, running mode, publishing, licensing, SDLC information, HCI information, user friendly, easy to configure, easy to train, software re-engineering, cyclic or repetitive, easy to learn and use user training.\n\nFrom the observed results (Table 2) we conclude that almost all of the applications have good implementations of their methodology (algorithms etc.) but often lack in usage point of views (user interface, documentation etc.), or long term sustainable development of the software (at least regarding the organization of future further developments). These shortcomings are prevented when following the “Butterfly” paradigm.\n\n\nConclusions\n\nIn the earlier sections of this paper we presented the concepts of usage of the existing scientific software solution design, modeling, implementation, testing and deployment. This helps in resolving conflicts and highlighting valuable differences between the traditional (professional) and scientific applications. This paper also proposes a new approach towards user friendly scientific software solution development, with emphases on the use of proper SDLC, HCI and technologies. The successful implementation of the five applications discussed strongly validates the potential of the Butterfly model.\n\nIn conclusion, although the adaptation of SSE principles to the Butterfly model may seem to increase developmental work load in comparison to the current running programming method applications, the Butterfly model will ultimately reduce the work by making the scientific application well designed, flexible, structured, reusable, developed according to a product line, as well as analytical and of high quality. According to its design, the software developed using the Butterfly paradigm is user friendly, easy to learn and deploy.", "appendix": "Author contributions\n\n\n\nZeeshan Ahmed has proposed the model Butterfly, and initiated the topic of discussion. Saman Zeeshan assisted Zeeshan Ahmed and Thomas Dandekar guided the study.\n\nAll authors participated in writing the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no conflict of interest.\n\n\nGrant information\n\nFunding was provided by the German Research Foundation (DFG), collaborative research center SFB 1047 \"Insect timing\", Project Z, to Zeeshan Ahmed. Thomas Dandekar and Saman Zeeshan were supported by the German Research Foundation (DFG), TR 34/Z.\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 all our interested colleagues for critical community input on the “Butterfly” approach.\n\nWe would like to thank Deutsche Forschungsgemeinschaft (DFG) for funding and University of Wuerzburg Germany for the support.\n\nWe thank two anonymous reviewers for helpful comments on the manuscript and the Land of Bavaria Germany.\n\n\nReferences\n\nde Champeaux D, Constantine L, Jacobson I, et al.: Structured analysis and object oriented analysis. In OOPSLA/ECOOP '90 Proceedings of the European conference on object-oriented programming on Object-oriented programming systems, languages, and applications. 1990. Reference Source\n\nAbrahamsson P, Salo O, Ronkainen J, et al.: Agile software development methods - Review and analysis. VTT Pub. 2002; 478. Reference Source\n\nManyika J, Chui M, Brown B, et al.: Big data: The next frontier for innovation, competition, and productivity. McKinsey Global Institute. 2011. Reference Source\n\nSergio C, Luciano F, Massimo B: Software Interoperability in consequence assessment: results of a feasibility study. Chem Eng Trans. 2010; 19: 341–346. Publisher Full Text\n\nBelhajjame K, Wolstencroft K, Corcho O, et al.: Metadata Management in the Taverna Workflow System. In IEEE International Symposium on Cluster Computing and the Grid. 2008. Publisher Full Text\n\nPenn State, Eberly College of Science. Galaxy DNA-analysis software is now available 'in the cloud'. ScienceDaily, Online 15 November 2011. Reference Source\n\nPabinger S, Dander A, Fischer M, et al.: A survey of tools for variant analysis of next-generation genome sequencing data. Brief Bioinform. First published online: January 21, 2013. PubMed Abstract \| Publisher Full Text\n\nMagrane M, Consortium U: UniProt Knowledgebase: a hub of integrated protein data. Database (Oxford). 2011: 2011; bar009. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoehm B: Software Engineering. IEEE Trans On Computers. 1976; 25(12): 1226–1242. Publisher Full Text\n\nRook P: Controlling software projects. Software Engin J. 1986; 1(1): 7–16. Publisher Full Text\n\nMahmood S, Lai R: RE-UML: A Component-Based System Requirements Analysis Language. Comput J. 2013; 56(7): 901–922. Publisher Full Text\n\nSzyperski C, Gruntz D, Murer S: Component Software: Beyond Object-Oriented Programming. Addison-Wesley. 2002. Reference Source\n\nBenediktsson O, Dalcher D, Thorbergsson H: Comparison of software development life cycles: a multiproject experiment. IEE Proceedings – Software. 2006; 153(3): 87–101. Publisher Full Text\n\nMunassar NMA, Govardhan A: A Comparison Between Five Models Of Software Engineering. Int Jr Comp Sci. 2010; 7(5): 94–101. Reference Source\n\nAhmed Z: Towards Performance Measurement and Metrics based Analysis of PLA Applications. Int J Software Engin App. 2010; 1(3): 66–80. Publisher Full Text\n\nAhmed Z, Majeed S: Measurement, Analysis with Visualization for better Reliability. In Artificial Intelligence and Hybrid Systems. C. Rocha, F. Akune, A. El-Shafie, iConcept Press, 2013. Reference Source\n\nAhmed Z, Majeed S: Towards Increase in Quality by Preprocessed Source Code and Measurement Analysis of Software Applications. IST Tran Inf Tech Theo App. 2010; 1(2): 8–13. Reference Source\n\nAhmed Z: Measurement Analysis and Fault Proneness Indication in Product Line Applications (PLA). In Sixth International Conference on New Software Methodologies, Tools, and Techniques. Italy, 2007; 391–400. Reference Source\n\nAhmed Z: Integration of variants handling in M-system NT. M.S. thesis, Blekinge Institute of Technology, Karlskrona, Sweden. 2006; 65. Reference Source\n\nLee J, Xue NL: Analyzing user requirements by use cases: a goal-driven approach. IEEE Softw. 1999; 16(4): 92–101. Publisher Full Text\n\nSommerville I: Integrated requirements engineering: a Tutorial. IEEE Softw. 2005; 22(1): 16–23. Publisher Full Text\n\nvan Lamsweerde A, Darimont R, Letier E: Managing conflicts in goal-driven requirements engineering. IEEE Trans Softw Eng. 1998; 24(11): 908–926. Publisher Full Text\n\nKaur H, Singh P: UML (Unified Modeling Language): Standard Language for Software Architecture Development. In International Symposium on Computing, Communication, and Control. Singapore, 2011. Reference Source\n\nGarlan D, Shaw M: An introduction to software architecture. In Advances in Software Engineering and Knowledge Engineering. V. Ambriola and G. Tortora: World Scientiﬁc Publishing Company, 1993; 2: 1–39. Reference Source\n\nGarlan D: Formal Approaches to Software Architecture. In Workshop on Studies of Software Design, UK 1993; 64–76. Reference Source\n\nGarlan D, Notkin D: Formalizing design spaces: Implicit invocation mechanisms. In 4th International Symposium of VDM Europe on Formal Software Development, UK 1991; 31–44. Reference Source\n\nDashofy EM, Hoek A, Taylor RN: An infrastructure for the rapid development of XML-based architecture description languages. In Twenty Fourth International Conference on Software Engineering, USA. 2002; 266–276. Publisher Full Text\n\nEgyed A, Kruchten PB: Rose/Architect: A Tool to Visualize Architecture. In Thirty Second Annual Hawaii Conference on Systems Sciences USA. 1999; 8: 8066. Reference Source\n\nBooch G, Rumbaugh J, Jacobson I: Unified Modeling Language User Guide, the (2nd Edition). Addison-Wesley Professional. 2005. Reference Source\n\nJacobson I, Christerson M, Jonsson P, et al.: Object-Oriented Software Engineering: A Use Case Driven Approach. Reading, MA: Addison-Wesley, 1992. Reference Source\n\nDumas M, ter-Hofstede AHM: UML Activity Diagrams as a Workflow Specification Language. In Fourth International Conference on The Unified Modeling Language, Modeling Languages. Concepts, and Tools, UK 2001; 2185: 76–90. Publisher Full Text\n\nBruza PD, van-der-Weide TP: The Semantics of Data Flow Diagrams. In International Conference on Management of Data. 1993. Reference Source\n\nLatronico E, Koopman P: Representing Embedded System Sequence Diagrams as a Formal Language. In Fourth International Conference on The Unified Modeling Language. Canada, 2001; 2185: 302–316. Publisher Full Text\n\nMarilyn B: A guide for programmers. Prentice-Hall, 1978. Reference Source\n\nBerardi D, Calvanese D, Giacomo GE: Reasoning on UML class diagrams. Artif Intell. 2005; 168(1–2): 70–118. Publisher Full Text\n\nHaney FM: Module connection analysis: a tool for scheduling of software debugging activities. Proceedings of Fall Joint Computer Conference. 1972; 173–179. Publisher Full Text\n\nMoreton R: A Process Model for Software Maintenance. Journal Information Technology. 1990; 5: 100–104. Publisher Full Text\n\nKan SH, Basili VR, Shapiro LN: Software Quality: An overview from the perspective of total quality management. IBM Systems Journal. 1994; 33(1). Reference Source\n\nLi W, Henry S: An Empirical Study of Maintenance Activities in Two Object-oriented Systems. Journal of Software Maintenance, Research and Practice. 1995; 7(2): 131–147. Publisher Full Text\n\nPfleeger SL, Bohner SA: A Framework for Software Maintenance Metrics. IEEE Transactions on Software Engineering. 1990; 320–327. Publisher Full Text\n\nMoreton R: A Process Model for Software Maintenance. Journal Information Technology. 1990; 5: 100–104. Publisher Full Text\n\nSoong NL: A program stability measure. In Proceedings of Annual ACM conference. Boulder Colorado, 1977; 163–173. Reference Source\n\nYau SS, Collofello JS, McGregor TM: Ripple effect analysis of software maintenance. In Proceedings COMPSAC '78. 1978; 60–65. Reference Source\n\nBlack S: Automating ripple effect measurement. In 5th World Multiconference on Systemics, Cybernetics and Informatics, Florida, USA. 2001. Reference Source\n\nDavis A: Software Requirements: Analysis and Specification. Prentice-Hall. New Jersey, 1989. Reference Source\n\nMartin J, McClure C: Software Maintenance: The Problem and its Solutions. Prentice-Hall. London, 1983. Reference Source\n\nParikh G: Some Tips, Techniques and Guidelines for Program and System Maintenance. Winthrup Publishers, Cambridge, Mass. 1982; 65–70. Reference Source\n\nSharpley WK: Software Maintenance Planning for Embedded Computer Systems. In Proceedings of the IEEE COMPSAC. 1977; 520–526. Reference Source\n\nOsborne WM: Building and Sustaining Software Maintainability. In Proceedings of Conference on Software Maintenance. 1987; 13–23. Reference Source\n\nYau SS, Collofello JS: Some Stability Measures for Software Maintenance. IEEE Trans On Software Engineering. 1980; 6(6): 545–552. Publisher Full Text\n\nJaffe MS, Leveson NG, Heimdahl MPE, et al.: Software requirements analysis for real-time process-control systems. IEEE Transactions on Software Engineering. 1991; 17(3): 241–258. Publisher Full Text\n\nChikofsky EJ, Cross JH: Reverse Engineering and Design Recovery: A Taxonomy. IEEE Soft. 1990; 7(1): 13–17. Publisher Full Text\n\nPetersen K, Wohlin C, Baca D: The Waterfall Model in Large-Scale Development. Product-Focused Software Process Improvement, Lecture Notes in Business Information Processing. 2009; 32: 386–400. Publisher Full Text\n\nRook P: Controlling software projects. Softw Eng J. 1986; 1: 7–16. Publisher Full Text\n\nBoehm BW: A spiral model of software development and enhancement. Computer. 1988; 21(5): 61–72. Publisher Full Text\n\nLarman C, Basili VR: Iterative and Incremental Development: A Brief History. Computer. 2003; 36(6): 47–56. Publisher Full Text\n\nHull C, Feygin M, Baron Y, et al.: Rapid prototyping: current technology and future potential. Rapid Prototyping Journal. 1995; 1(1): 11–19. Publisher Full Text\n\nAmbler S: Agile Modeling: Effective Practices for eXtreme Programming and the Unified Process. Wiley Computer Publishing. 2002. Reference Source\n\nCheriet H, Bounour N: Software evolution: Models and challenges. In International Conference on Machine and Web Intelligence (ICMWI), 2010; 479–481. Publisher Full Text\n\nPei Y, Wei Y, Furia CA, et al.: Code-Based Automated Program Fixing. In 26th IEEE/ACM International Conference on Automated Software Engineering (ASE). 2011; 392–395. Publisher Full Text\n\nWilliam K: A simple guide to five normal forms in relational database theory. Commun ACM. 1983; 26(2): 120–125. Publisher Full Text\n\nCodd EF: Normalized data base structure: A brief tutorial. In ACM SIG-FIDET Workshop on Data Description, Access and Control, San Diego, 1971; 1–17. Publisher Full Text\n\nCodd EF: Further normalization of the data base relational model. IBM Res Rep. 1971; RJ909. Reference Source\n\nFagin R: Multivalued dependencies and a new normal form for relational databases. ACM Trans on Database Sys. 1977; 2(3): 262–278. Publisher Full Text\n\nFagin R: Normal forms and relational database operators. In ACM SIGMOD International Conference on Management of Data, USA. 1979; 153–160. Publisher Full Text\n\nAhmed Z, Ganti SK, Kyhlbäck H: Design Artifact’s, Design Principles, Problems, Goals and Importance. In Fourth International Conference of Statistical Sciences, Pakistan. 2008; 15: 57–68. Reference Source\n\nAhmed Z: Designing Flexible GUI to Increase the Acceptance Rate of Product Data Management Systems in Industry. Int J Comp Sci Emerg Tech. 2011; 2(1): 100–109. Reference Source\n\nKlemmer SR, Lee B: Notebooks that Share and Walls that Remember: Electronic Capture of Design Education Artifacts. In ACM Symposium on User Interface Software and Technology. 2005. Reference Source\n\nAhmed Z, Zeeshan S, Huber C, et al.: Software LS-MIDA for efficient mass isotopomer distribution analysis in metabolic modelling. BMC Bioinformatics. 2013; 14: 218. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAhmed Z, Majeed S, Dandekar T: Unified Modeling and HCI Mockup Designing towards MIDA. Int Jr Emerg Sci. 2012; 2(3): 361–382. Reference Source\n\nAhmed Z, Helfrich-Förster C, Dandekar T: Integrating Formal UML Designs and HCI Patterns with Spiral SDLC in DroLIGHT Implementation. Rec Pat Comp Sci. 2013; 6(2): 85–98. Publisher Full Text\n\nAhmed Z, Helfrich-Förster C: DroLIGHT: Real Time Embedded System towards Endogenous Clock Synchronization of Drosophila. Front Neuroinform Conference Abstract: Neuroinformatics. 2013. Publisher Full Text\n\nAhmed Z, Helfrich-Förster C: DroLIGHT-2: Real Time Embedded and Data Management System for Synchronizing Circadian Clock to the Light-Dark Cycles. Rec Pat Comp Sci. 2013; 6(3): 191–205. Publisher Full Text\n\nAhmed Z, Majeed S, Dandekar T: Formal UML Modelling of Isotopo, Bioinformatical Software for Mass Isotopomers Distribution Analysis. Software Engin. 2012; 2: 147–159. Publisher Full Text\n\nWiechert W, de Graaf AA: Bidirectional reaction steps in metabolic networks: I. Modeling and simulation of carbon isotope labeling experiments. Biotechnol Bioeng. 1997; 55(1): 101–117. PubMed Abstract \| Publisher Full Text\n\nSchuster R, Schuster S: Refined algorithm and computer program for calculating all non-negative fluxes admissible in steady states of biochemical reaction systems with or without some flux rates fixed. Comput Appl Biosci. 1993; 9(1): 79–85. PubMed Abstract \| Publisher Full Text\n\nCvijovic M, Olivares-Hernández R, Agren R, et al.: BioMet Toolbox: genome-wide analysis of metabolism. Nucleic Acids Res. 2010; 38(Web Server issue): 144–149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZamboni N, Fischer E, Sauer U: FiatFlux - a software for metabolic flux analysis from 13C-glucose experiments. BMC Bioinformatics. 2005; 6: 209. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPitkänen E, Akerlund A, Rantanen A, et al.: ReMatch: a web-based tool to construct, store and share stoichiometric metabolic models with carbon maps for metabolic flux analysis. J Integr Bioinformatics. 2008; 5(2): 1–13. PubMed Abstract \| Publisher Full Text\n\nKlamt S, von Kamp A: An application programming interface for CellNetAnalyzer. Biosystems. 2011; 105(2): 162–168. PubMed Abstract \| Publisher Full Text\n\nAhmed Z, Majeed S, Dandekar T: Computational Feature Performance and Domain Specific Architecture Evaluation of Software Applications Towards Metabolic Flux Analysis. Rec Pat Comp Sci. 2012; 5(3): 165–176. Publisher Full Text\n\nDandekar T, Fieselmann A, Majeed S, et al.: Software applications toward quantitative metabolic flux analysis and modeling. Brief Bioinform. 2014; 15(1): 91–107. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4632", "date": "12 May 2014", "name": "Paul Vauterin", "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 addresses some major and often overlooked problems associated with software development in an academic environment:Lack of long-term vision due to short contract cycles.Lack of attention to usability aspects such as well-thought user interfaces.Lack of documentation, training material, and other follow-up.I share the opinion of the authors that these elements pose a substantial threat to the quality of the scientific software products that are created in academic environment. Lots of valuable academic software development projects fail because of one or more of these reasons, resulting in a waste of efforts. Therefore, I think that papers focusing on solutions to overcome these issues certainly deserve a place. A substantial portion of the paper reads like an introduction to standard software engineering practices, with a focus on scientific software development. It is generally well written, and does provide a useful resource for academic software developers who are interested in learning more about this subject. In addition, the authors present the “Butterfly model”, which they claim is a novel approach to software engineering, specifically tailored towards scientific software development. It does contain some interesting aspects, specifically the integrated attention to Human-Computer-Interaction (HCI), something often vastly overlooked in academic software development. However, in my opinion the authors tend to overstate the originality of this approach, as it mostly recuperates already known concepts and practices. It does have the merit that it attempts to capture these concepts in a single framework focusing on scientific software development in an academic environment. Perhaps the major problem I have with the way this model is presented is that the authors do not clearly mention its scope of application. Clearly, this model only makes sense for larger and longer-term development projects, whereas lots of academic (bioinformatics) software projects are very small, simple in structure, and short-lived by nature. Applying the “Butterfly model” or a similar rather heavyweight approach to these projects could be a counter-productive and frustrating. In my opinion, the authors missed the opportunity to stratify the preferred approach according to the size and complexity of the project at hand. A consequent risk is that a reader might be confused and wrongly informed, thinking that this approach should be used for every software development effort, regardless its size. I think that the authors should properly address the scope of the model in the paper. The authors provide a rather large set of software products they claim were developed with the Butterfly model, but do not provide much detail about what this meant in practice. I believe that the paper would vastly benefit if one specific case was picked, for which the practical approach of the Butterfly model would be elaborated, and the benefits would be highlighted in a very concrete way. This would also help to substantiate the author's claim about the novelty of the Butterfly model. Further comments: “Moreover the SDLCs famous for the quick development (Rapid Prototype Model, Agile Development Model, Extreme Programming Model, Evolutionary Model, Code and Fix Model), lack in quality software production”:This is a rather controversial statement, as the list includes some generally respected models. Since the paper does not provide further arguments to back this claim, I believe it should not be made. “Here, function requirements are those which can be implemented (based on existing resources, time, budget, labor, tools, technologies and methodologies), and non-functional requirements are those which cannot be implemented”:I do not believe that this the standard definition of function(al) vs. non-functional requirements. Functional requirements specify what the system should do (e.g. input-behaviour-output), whereas non-functional requirements specify how it should do that (e.g. qualities such as speed, robustness, scalability, maintainability, …). “Real time examples using Butterfly”: I think this should read “Real life examples…”. Other ways, I can’t interpret this title. The link to the “Isotopo software” (http://www.tr34.uni-wuerzburg.de/computations/isotopo/) was broken at the time this report was written. In an ironic way, this seems to illustrate one of the points of the authors about the short-livedness of many academic software projects. I believe that this article should be catalogued as “Methods Article” rather than “Research article”.", "responses": [ { "c_id": "924", "date": "27 Jul 2014", "name": "Zeeshan Ahmed", "role": "Author Response", "response": "Thank you so much for your time in reviewing our manuscript and giving valuable suggestions, which all helped to further improve our manuscript. Referee Comment:\"This paper addresses some major and often overlooked problems associated with software development in an academic environment:Lack of long-term vision due to short contract cycles.Lack of attention to usability aspects such as well-thought user interfaces.Lack of documentation, training material, and other follow-up.\" We agree with you. Referee Comment:\"I share the opinion of the authors that these elements pose a substantial threat to the quality of the scientific software products that are created in academic environment. Lots of valuable academic software development projects fail because of one or more of these reasons, resulting in a waste of efforts. Therefore, I think that papers focusing on solutions to overcome these issues certainly deserve a place.\" Thanks. Referee Comment:\"A substantial portion of the paper reads like an introduction to standard software engineering practices, with a focus on scientific software development. It is generally well written, and does provide a useful resource for academic software developers who are interested in learning more about this subject.\" We agree with you, and thanks for appreciating these points. Referee Comment:\"In addition, the authors present the “Butterfly model”, which they claim is a novel approach to software engineering, specifically tailored towards scientific software development. It does contain some interesting aspects, specifically the integrated attention to Human-Computer-Interaction (HCI), something often vastly overlooked in academic software development. However, in my opinion the authors tend to overstate the originality of this approach, as it mostly recuperates already known concepts and practices. It does have the merit that it attempts to capture these concepts in a single framework focusing on scientific software development in an academic environment.\" We agree with you, stressing important points of our paper.Referee Comment:\"Perhaps the major problem I have with the way this model is presented is that the authors do not clearly mention its scope of application. Clearly, this model only makes sense for larger and longer-term development projects, whereas lots of academic (bioinformatics) software projects are very small, simple in structure, and short-lived by nature. Applying the “Butterfly model” or a similar rather heavyweight approach to these projects could be counter-productive and frustrating. In my opinion, the authors missed the opportunity to stratify the preferred approach according to the size and complexity of the project at hand. A consequent risk is that a reader might be confused and wrongly informed, thinking that this approach should be used for every software development effort, regardless its size. I think that the authors should properly address the scope of the model in the paper.The authors provide a rather large set of software products they claim were developed with the Butterfly model, but do not provide much detail about what this meant in practice. I believe that the paper would vastly benefit if one specific case was picked, for which the practical approach of the Butterfly model would be elaborated, and the benefits would be highlighted in a very concrete way. This would also help to substantiate the author's claim about the novelty of the Butterfly model.\"Thanks for these well taken suggestions - we agree with you, and tried to incorporate these. We have heavily revised the manuscript and now, along with some previously given examples, we have added a case study. Each layer of the Butterfly model is now explained with this example of its implementation, explaining the way we have applied the concepts in a real time scientific software solution’s development. We hope this can clarify most of the ambiguous or unclear points including getting an idea on the specific merits and originality as well as the stratification of the approach, e.g. how much effort should be devoted at each step, looking at the concrete examples given.Referee Comment:\"Further comments: “Moreover the SDLCs famous for the quick development (Rapid Prototype Model, Agile Development Model, Extreme Programming Model, Evolutionary Model, Code and Fix Model), lack in quality software production”:This is a rather controversial statement, as the list includes some generally respected models. Since the paper does not provide further arguments to back this claim, I believe it should not be made.\"Following your suggestion, we have revised this paragraph accordingly. Referee Comment:\"“Here, function requirements are those which can be implemented (based on existing resources, time, budget, labor, tools, technologies and methodologies), and non-functional requirements are those which cannot be implemented”:I do not believe that this the standard definition of function(al) vs. non-functional requirements. Functional requirements specify what the system should do (e.g. input-behaviour-output), whereas non-functional requirements specify how it should do that (e.g. qualities such as speed, robustness, scalability, maintainability).\"We agree and thank you for the clarification, the paragraph has been corrected accordingly. Referee Comment:“Real time examples using Butterfly”: I think this should read “Real life examples…”. Other ways, I can’t interpret this title.Following your suggestion, we have revised this sentence. Referee Comment:\"The link to the “Isotopo software” (http://www.tr34.uni-wuerzburg.de/computations/isotopo/) was broken at the time this report was written. In an ironic way, this seems to illustrate one of the points of the authors about the short-livedness of many academic software projects.We have revised and updated the web link.Referee Comment:\"I believe that this article should be catalogued as “Methods Article” rather than “Research article”.\"We agree with you, and we changed the category of the manuscript.Thank you so much for your valuable time in reading and evaluating our work, giving your advice to further improve it. We tried to follow your suggestions as much as possible in the revision. Furthermore, all those questions which are still incompletely resolved due to the scope or limitations of this paper, we will work on further in the future." } ] }, { "id": "5147", "date": "09 Jul 2014", "name": "Wolfgang Mueller", "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\nCaveat: I lead a group that does scientific software development. I am interested in software development both from the conceptual and the technical side. However, I would not be a serious contender for a software engineering chair. As a consequence, I cannot claim full knowledge about the state of the art of software engineering processes.I agree to the following takeaways that I pull out of the paper:If you know you are building an application which is for other people than yourself, it makes sense to follow a software engineering process.Think about maintainability and sustainability of your software.Think about your users and write usable software.Operate using cyclic refinementThe Butterfly model is an assembly of current software practices, emphasizing the needs of the scientific software developer. I see the strength of the article in giving a concise overview of what a scientific software developer should do.However, I disagree with some statements made in the paper. Some of these statements relate rather to anecdotal experience. This points to one of the weaknesses in the paper: There are statements made with respect to the scientific software domain that are not marked as anecdotal, but which are not backed by either empirical facts (e.g. questionnaires) or pointers to such facts in the literature.In more detail:Generally I agree, science is a fast-paced environment with rapidly changing requirements asking for suitable software engineering processes.Current software engineering and developmentThe paper comprises a large number of citations. However, I do not agree with the insights taken by the authors from that literature.To my knowledge, the field of Software Engineering has realized that one key factors for success is if software matches the needs of users. Agile Development follows the view that often-times, users don't even know what they need until they are closely involved in development and see where their own ideas of their work (i.e. using the program) could fail. This is why there is less of an initial requirements analysis, but there is a series of sprints that each seek the implementation of features. Each feature implemented also influences the next stage of the requirements analysis.The authors of this paper see the lack of quality software production as one of the weaknesses of agile software engineering practices. However, Agile software engineering practices like XP and Scrum emphasize the importance of testing and other practices intended to increase the software quality. Such processes definitely do not disregard the production of quality production. So, to cut a long story short, I do not agree with the impression created in the paper that agile processes lead to muddle-through software that is unstable and short term due to the process.You criticize the \"ripples\" through the program that come from changes. Ripples is the word describing the needs for software changes due to one change elsewhere in the program e.g. the change in the view necessitated by changing a model. The paper defining ripples is from 1978, and you cite another paper about measuring ripples (2001). However to my knowledge, in agile development, refactoring (improving code quality without changing its semantic meaning) is seen as an effort that accompanies development, with a view on minimizing ripples. Unit, integration, acceptance tests are automated as far as possible so they can be run after almost every change. Ongoing refactoring without being able to test easily if the program still works is admittedly very hard. But tests well done can help with this. I think it would be useful for the reader if you would outline how ripples are relevant in XP or Scrum.The fact that user interaction is not written into the Scrum process does not mean that it's not on the map. It is implicit in the requirements given by the users and users being part of the Agile team. Typically, users that are part of the Agile team will accept or not the software created. User tests can be part of the acceptance procedure, usability improvements can be features in a Scrum process.While user experience is put to the forefront, the paper is quite sparse on the processes and on citations regarding HCI.You state: \"Unfortunately HCI is the most ignored and unattended phase of scientific software solution development\". To my impression, this is not the state of the art any more, and the EBI even have specialized UX personnel who can be called into projects.The authors then describe what they see as reasons for poor HCI properties of scientific software and state that no-one uses software with bad HCI and state that software with bad HCI won't be used. In my (anecdotal) experience, many software products start out as concrete problem solvers for their authors, with bad HCI (for others, as the original author writes the software tuned to his/her own needs) which then are successively refined to serve a larger public. Of course this process bears its problems for outside users. However, such software gets adopted, because of word of mouth.In the article, HCI \"design patterns\" are not laid out in sufficient detail. The normal forms (1NF) etc. could be either treated more shortly or in more detail. The enumerating of 1NF to 5NF and not saying what they signify is in my view not the right middle ground.---\"Real time examples\" should be changed to \"Real life examples\"I think these examples are an interesting overview, but to my feeling they are not addressing what is needed to support the claims made about the Butterfly process. Rather than giving a list of software artifacts that were created using the process I would prefer more information about the outcome of the different phases. What were iterative improvements etc.? If you are using a ticketing systems and a versioning system, it is highly likely that you can generate from logs a compelling story that lets the readers know more about the Butterfly process in everyday life. In particular it would be interesting to learn more about how many people performed the processes, and how the roles were distributed. Such an approach would also benefit the software system comparison of systems that did not use Butterfly. Following the Butterfly paradigm will take time and effort. Was a sufficient amount of time available for the systems used for the overview?---All in all, I find this paper is a mixture between an opinion piece, a methods piece, and an overview piece. Even if I do not agree with many claims made in the paper, I do think the question of whether scientific software engineering (or even some scientific domains) need special software engineering processes is an interesting one (albeit maybe not a new one).", "responses": [ { "c_id": "923", "date": "27 Jul 2014", "name": "Zeeshan Ahmed", "role": "Author Response", "response": "Thank you so much for your time in reviewing our manuscript and giving valuable suggestions, as we believe that following those our manuscript has been improve a level ahead.Referee Comment:\"Caveat: I lead a group that does scientific software development. I am interested in software development both from the conceptual and the technical side. However, I would not be a serious contender for a software engineering chair. As a consequence, I cannot claim full knowledge about the state of the art of software engineering processes.I agree to the following takeaways that I pull out of the paper: If you know you are building an application which is for other people than yourself, it makes sense to follow a software engineering process.Think about maintainability and sustainability of your software.Think about your users and write usable software.Operate using cyclic refinement We agree with you.Referee Comment:\"The Butterfly model is an assembly of current software practices, emphasizing the needs of the scientific software developer. I see the strength of the article in giving a concise overview of what a scientific software developer should do.\" Thanks, and we agree with you.Referee Comment:\"However, I disagree with some statements made in the paper. Some of these statements relate rather to anecdotal experience. This points to one of the weaknesses in the paper: There are statements made with respect to the scientific software domain that are not marked as anecdotal, but which are not backed by either empirical facts (e.g. questionnaires) or pointers to such facts in the literature.\" Thank you for pointing this out - we respect your opinion.Referee Comment:\"In more detail: Generally I agree, science is a fast-paced environment with rapidly changing requirements asking for suitable software engineering processes.\" True.Referee Comment:\"Current software engineering and development The paper comprises a large number of citations. However, I do not agree with the insights taken by the authors from that literature.\" Thanks, and we respect your opinion. Referee Comment:\"To my knowledge, the field of Software Engineering has realized that one key factors for success is if software matches the needs of users.\" True. Referee Comment:\"Agile Development follows the view that often-times, users don't even know what they need until they are closely involved in development and see where their own ideas of their work (i.e. using the program) could fail. This is why there is less of an initial requirements analysis, but there is a series of sprints that each seek the implementation of features. Each feature implemented also influences the next stage of the requirements analysis.\" We agree with you.Referee Comment:\"The authors of this paper see the lack of quality software production as one of the weaknesses of agile software engineering practices. However, Agile software engineering practices like XP and Scrum emphasize the importance of testing and other practices intended to increase the software quality. Such processes definitely do not disregard the production of quality production. So, to cut a long story short, I do not agree with the impression created in the paper that agile processes lead to muddle-through software that is unstable and short term due to the process.\" We respect your opinion and want to explain here that the aim of this paper is not at all to present Agile software engineering practices as muddle-through approaches, but to state that the “Butterfly” paradigm is more useful in some specific scenarios. Our aim is to generalize the basic software engineering concepts and stream line them in a way that most of the people in Bioinformatics (non-informatics or with more biology background) can easily adopt them. Hence, Agile software engineering practice is of course a viable approach, and we have revised our paper and tried to elaborate the proposed new approach in more detail, by giving a case study of a real time scientific solution’s development and being careful not to misrepresent alternatives.Referee Comment:\"You criticize the \"ripples\" through the program that come from changes. Ripples is the word describing the needs for software changes due to one change elsewhere in the program e.g. the change in the view necessitated by changing a model. The paper defining ripples is from 1978, and you cite another paper about measuring ripples (2001). However to my knowledge, in agile development, refactoring (improving code quality without changing its semantic meaning) is seen as an effort that accompanies development, with a view on minimizing ripples.\" Thanks for addressing “ripples”. We respect your opinion and you’re points are valid, in particular regarding Agile development. However, our point is summarised here (expanded on in revised text):Adding new features to existing applications on a random basis in academia, can cause unidentified and illogical errors to occur which can badly effect the stability of the application. For example - we have implemented some operations with mutual consent of all the team members (including scientists, students, lab operators and computer scientists etc.) and then when the application’s new version is completed, a new change comes from the team. This requires the selective removal of implemented options with the addition of new connecting features, giving a high probability that errors will occur. Referee Comment:\"Unit, integration, acceptance tests are automated as far as possible so they can be run after almost every change. Ongoing refactoring without being able to test easily if the program still works is admittedly very hard. But tests well done can help with this. I think it would be useful for the reader if you would outline how ripples are relevant in XP or Scrum.\" We respect your opinion and you are absolutely right. So we now touch upon the questions of ripples regarding XP and scrum. In our personal academic experience we have noticed that especially in those cases when the developer is not the user of the real time scientific software application (which will be used in the laboratories or in fields e.g. our developed solutions DroLIGHT or Ant-App-DB), it is really hard to perform extensive on-time testing (including Unit, integration, acceptance tests etc.). One cannot receive the greatest benefit from testing in these situations. So when a developer has to keep randomly and extensively changing things in an application without on-time testing then there is a high probability of getting ripple effects.Referee Comment:\"The fact that user interaction is not written into the Scrum process does not mean that it's not on the map. It is implicit in the requirements given by the users and users being part of the Agile team. Typically, users that are part of the Agile team will accept or not the software created. User tests can be part of the acceptance procedure, usability improvements can be features in a Scrum process. While user experience is put to the forefront, the paper is quite sparse on the processes and on citations regarding HCI.You state: \"Unfortunately HCI is the most ignored and unattended phase of scientific software solution development\". To my impression, this is not the state of the art any more, and the EBI even have specialized UX personnel who can be called into projects.\"You are absolutely right and we agree with you but this is more applicable when a commercial application is developed or some specialised software engineering or related group is producing the solution. We added HCI citations and mention the EBI approach as an example of recent trends. Furthermore, it is true, HCI is extensively ignored when a new scientific software solutions is developed in most of the scientific academia.Along with the heavy scientific software solution development, we are also taking part in analysing different scientific software applications for different purposes e.g. Dandekar T, Fieselmann A, Saman M, and Zeeshan Ahmed. \"Software Applications toward Quantitative Metabolic Flux Analysis and Visualization\", Briefings in Bioinformatics, Oxford University Press, Oxford Journals; 15(1): 91-107, 2014.Zeeshan Ahmed, Saman M, Dandekar T: Computational feature performance and domain specific architecture evaluation of software applications towards metabolic flux analysis. Recent Patents on Computer Science, 5(3):165-176, 2012. We have analysed and used many different developed software applications (more than 35) and scripts. With reference to our published work, most of the time is consumed in configuring the applications and understanding its features. If a person with an Informatics background cannot easily configure and use Bioinformatics solutions then how can a layman.Moreover, when it comes to follow some already developed open source application, and add some features to it - then if the pre-processed source code is quite large and without any formal documentation, it can be a nightmare for Bioinformatics students.Referee Comment:\"The authors then describe what they see as reasons for poor HCI properties of scientific software and state that no-one uses software with bad HCI and state that software with bad HCI won't be used. In my (anecdotal) experience, many software products start out as concrete problem solvers for their authors, with bad HCI (for others, as the original author writes the software tuned to his/her own needs) which then are successively refined to serve a larger public. Of course this process bears its problems for outside users. However, such software gets adopted, because of word of mouth.\" We respect your opinion and agree with you, and have altered the statements accordingly the revision.Referee Comment:\"In the article, HCI \"design patterns\" are not laid out in sufficient detail.\" Following your suggestion, we have revised it and have added some information about HCI design patterns.Referee Comment:\"The normal forms (1NF) etc. could be either treated more shortly or in more detail. The enumerating of 1NF to 5NF and not saying what they signify is in my view not the right middle ground.\" Thanks for the opinion and following your suggestion, we have revised it. Here our point was not to elaborate on database normalisation forms but emphasise its use. In our previous experience (without denigrating any researcher or developer), we have noticed that often, databases are created to store and manage data - but there are no relationships between entities. This can have a negative impact - especially in search and indexing operations. Moreover if data is well normalized then for large datasets, it will expedite the processing speed in searching the elements.Referee Comment:\"Real time examples\" should be changed to \"Real life examples\"Following your suggestion, we have revised the heading and the examples have been augmented.Referee Comment:\"I think these examples are an interesting overview, but to my feeling they are not addressing what is needed to support the claims made about the Butterfly process. Rather than giving a list of software artifacts that were created using the process I would prefer more information about the outcome of the different phases. What were iterative improvements etc.? If you are using a ticketing systems and a versioning system, it is highly likely that you can generate from logs a compelling story that lets the readers know more about the Butterfly process in everyday life. In particular it would be interesting to learn more about how many people performed the processes, and how the roles were distributed.\" We agree with you. We have revised our paper and tried to elaborate the proposed approach in more detail, by giving a case study of a real time scientific solution’s development. Questions such as iterative improvement are answered by concrete examples.Referee Comment:\"Such an approach would also benefit the software system comparison of systems that did not use Butterfly. Following the Butterfly paradigm will take time and effort. Was a sufficient amount of time available for the systems used for the overview?\" You are absolutely right here that there must be some comparison drawn between the software applications developed using Butterfly and without. The problem here is that our proposed paradigm is quite new and this paper is the first public, methodological debate on it. Moreover according to the scope of this manuscript, we have only presented our approach with some examples but in future we will definitely try to come up with more intensive comparisons. We are still working on the strengthening our approach. In particular, for sustainable development of an application the Butterfly paradigm did boost results - for instance developing metabolic modelling software (LS-MIDA, Isotopo, LipidPRO including database applications and data management) or neurobiology software (the running example of the paper, the DroLIGHT series of programs). We now present the approach to others and note that we also have good results in developing the first version of the App-Ant database, an application oriented behaviour monitoring system, again with a planned longer term pipeline development.Referee Comment:\"--- All in all, I find this paper is a mixture between an opinion piece, a methods piece, and an overview piece. Even if I do not agree with many claims made in the paper, I do think the question of whether scientific software engineering (or even some scientific domains) need special software engineering processes is an interesting one (albeit maybe not a new one).\" Thank you so much for your valuable time in reading and evaluating our work and giving your opinions to further improve it. We have tried to follow all your helpful suggestions and those which are still left partly open due to the scope or limits of this paper will soon be answered by further efforts. We have strengthened the methods section (see also our response to Paul Vauterin) but kept our opinion and own approach as well as added more citations for an overview." } ] } ]	1	https://f1000research.com/articles/3-71
https://f1000research.com/articles/3-181/v1	01 Aug 14	{ "type": "Case Report", "title": "Case Report: Solitary mastocytoma treated successfully with topical tacrolimus", "authors": [ "M. S. Sukesh", "Ameet Dandale", "Rachita Dhurat", "Ankur Sarkate", "Smita Ghate", "Ameet Dandale", "Rachita Dhurat", "Ankur Sarkate", "Smita Ghate" ], "abstract": "Solitary mastocytoma, a rare dermatological entity accounts for 10-15% of cutaneous mastocytosis. We report a rare case of solitary mastocytoma presenting at birth, treated successfully with topical tacrolimus. Along with reassurance and strict avoidance of triggering factors, no recurrence was reported within the one year follow-up period.", "keywords": [ "Solitary mastocytoma", "a rare dermatological entity", "represents the second most common type of cutaneous mastocytoma. Solitary mastocytomas constitute 10–20% of all childhood cutaneous mastocytosis. They usually present within 2 years of age", "mostly within first 3 months1." ], "content": "Introduction\n\nSolitary mastocytoma, a rare dermatological entity, represents the second most common type of cutaneous mastocytoma. Solitary mastocytomas constitute 10–20% of all childhood cutaneous mastocytosis. They usually present within 2 years of age, mostly within first 3 months1.\n\nWe report a case of solitary mastocytoma presenting a birth that was treated successfully with topical tacrolimus with no recurrences noted during a one year follow-up period.\n\n\nCase report\n\nAn eighteen month old girl presented with a solitary, itchy dark coloured, minimally elevated lesion over her left elbow that had been evident since birth. The lesion used to itch and swell on scratching, bathing and toweling of the area. The child was otherwise healthy and no other systemic manifestations were noted. Clinical examination revealed a solitary, 3.5 × 6.5 cm, non-tender, minimally elevated plaque with central shiny skin and peripheral marginal hyperpigmentation over left elbow. On scratching the lesion with the blunt end of a pin, the central shiny skin became edematous and itchy (positive Darier’s sign) (Figure 1). Hematological and biochemical investigations were within normal limits. A 5 mm biopsy of the skin tissue obtained from the center of the lesion revealed a dense monomorphic inflammatory infiltrate consisting of round to oval cells with clear cytoplasm and centrally located nuclei in the upper and mid dermis (Figure 2a, 2b). Special staining with toluidine blue revealed metachromatic staining of the monomorphic mast cells, confirming the diagnosis of mastocytoma (Figure 3).\n\na, Dense monomorphic inflammatory infiltrate in upper and mid dermis; b, Dense monomorphic inflammatory infiltrate consisting of round to oval cells with clear cytoplasm noted at 40× magnification.\n\nThe child was treated with topical tacrolimus 0.03% ointment which was applied on the lesion site twice daily. The child was also prescribed an oral antihistamine (levocetirizine syrup, 1.25 mg once a day). By the end of third month, complete subsidence of the lesion was noticed with residual hyperpigmentation, negative Darier's sign, and no signs of atrophy. This treatment was continued for another four months which led to resolution of the lesion with residual hyperpigmentation, negative Darier’s sign, and no signs of atrophy. Treatment was continued with only a once a day application of topical tacrolimus for a month after clinical resolution to prevent further recurrence (Figure 4). Reassurance and strict avoidance of triggering factors such as pressure, friction (rubbing or toweling of the lesion), extreme temperature changes, intake of mast cell degranulating agents like aspirin, NSAIDS, morphine, codeine (especially in the form of cough preparations) has led to no recurrence of the child’s symptoms during a 1 year follow-up period.\n\nThe central atrophic scar due to biopsy can be seen in the centre of the lesion.\n\n\nDiscussion\n\nSolitary mastocytoma, the second most common type of cutaneous mastocytosis, accounts for 10–15% of cutaneous mastocytosis1. Nearly half of solitary mastocytomas present within the first 3 months of life and the remaining half during the first year2. Solitary mastocytoma presenting in adults has also been noted3. The most common locations of mastocytomas are on the trunk, neck, and arms.\n\nMost solitary mastocytomas are about 1–5 cm in diameter and are seen as skin areas that are colored yellow to brown and present as minimally elevated plaques with a smooth shiny surface having a soft to rubbery consistency. The lesion turns edematous and itchy on manipulation [rubbing or trauma to the lesion]. Mild tenderness and the formation of vesicles or bulla can also occur4. These features can sometimes be so mild that they may not come to the attention of parents.\n\nDiagnosis is by biopsy that reveals a dense monomorphic inflammatory infiltrate consisting of round to oval mast cells containing a clear cytoplasm and centrally located nuclei in the dermis. Confirmation of diagnosis is usually by special staining with toluidine blue that reveals the metachromatic staining of the monomorphic mast cells5.\n\nThe course of solitary mastocytomas is benign and the disease is self-limited. Systemic involvement is uncommon and complete spontaneous resolution is expected in months to years’ time. Reassurance along with avoidance of triggering factors such as pressure, friction (rubbing or toweling of the lesion), physical exertion, extreme temperature changes, emotional stress, intake of mast cell degranulating agents like aspirin, NSAIDS, morphine, codeine (particularly in cough preparations), alcohol and radio contrast dyes are of utmost importance6.\n\nIn symptomatic patients, oral H1 and H2 antihistamines are commonly used. Topical steroids with or without occlusion, intralesional steroids, oral sodium cromoglycate, oral ketotifen and surgical excision are other treatment options6,7. Though topical steroids have shown good results, their topical and systemic side effects are a matter of concern, especially when treating infants.\n\nTacrolimus and pimecrolimus are topical immunomodulators, the first in a new class of topical calcineurin inhibitors. These drugs act as immunosuppressants by binding to a cytosolic ligand in the cytoplasm of T cells called FK506-binding protein (FKBP) and inhibit the cytoplasmic enzyme calcineurin, thus inhibiting the activation and maturation of T cells and blocking transcriptional activation of several cytokine genes – interleukin (IL)-2 [mainly], IL-4, IL-10, interferon-γ, tumor necrosis factor-α, and granulocyte–macrophage colony-stimulating factor8.\n\nOther immunomodulatory effects of tacrolimus include the inhibition of mast cell adhesion and the inhibition of the release of mediators from mast cells and basophils9, which might explain its efficacy in the improvement of the lesion and alleviation of the symptoms in cutaneous mastocytosis.\n\nThese immunomodulators offer advantages over corticosteroids in terms of a more selective action, no associated systemic side-effects, and the absence of associated skin atrophy, depigmentation and telangiectasia.\n\nThis case report demonstrates that topical calcineurin inhibitors can be considered as a safe and efficacious modality of treatment in cutaneous mastocytoma.\n\n\nConsent\n\nWritten informed consent for publication of the clinical details and clinical images was obtained from the father of the patient.", "appendix": "Author contributions\n\n\n\nDr. Sukesh M.S. and Dr. Ameet Dandale were involved in clinical diagnosis, work-up, treatment and writing up of this case report. Dr Smita Ghate contributed to the histopathologic diagnosis, Dr Rachita Dhurat contributed to the conception and design and final approval of the paper; Dr Ankur Sarkate contributed to the assimilation of all data and the histopathological pictures.\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\nVan Gysel D, Van Schail R, Oranje A: Mastocytosis. In Textbook Of Pediatric Dermatology (Harper J, Oranje A, and Prose N). 2nd Ed. Oxford: Blackwell Scientific. 2006; 703–717. Reference Source\n\nAkoglu G, Erkin G, Cakir B, et al.: Cutaneous mastocytosis: demographic aspects and clinical features of 55 patients. J Eur Acad Dermatol Venereol. 2006; 20(8): 969–973. PubMed Abstract \| Publisher Full Text\n\nKis Pandhi D, Singal A, Aggarwal S: Adult onset, hypopigmented solitary mastocytoma: report of two cases. Indian J Dermatol Venereol Leprol. 2008; 74(1): 41–3. PubMed Abstract \| Publisher Full Text\n\nChang J, Yang C, Sung F, et al.: A red-brown plaque on the nape. Solitary mastocytoma. Arch Dermatol. 2004; 140(10): 1275–1280. PubMed Abstract \| Publisher Full Text\n\nSridharan G, Shankar AA: Toluidine blue: A review of its chemistry and clinical utility. J Oral Maxillofac Pathol. 2012; 16(2): 251–255. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeide R, Beishuizen A, De Groot H, et al., Dutch National Mastocytosis Work Group: Mastocytosis in children: a protocol for management. Pediatr Dermatol. 2008; 25(4): 493–500. PubMed Abstract \| Publisher Full Text\n\nFlageul B: Cutaneous mastocytosis. Rev Prat. 2006; 56(16): 1745–1751. PubMed Abstract\n\nPrucha H, Schnopp C, Akdis C, et al.: Pimecrolimus, a topical calcineurin inhibitor used in the treatment of atopic eczema. Expert Opin Drug Metab Toxicol. 2013; 9(11): 1507–16. PubMed Abstract \| Publisher Full Text\n\nNghiem P, Pearson G, Langley RG: Tacrolimus and pimecrolimus: from clever prokaryotes to inhibiting calcineurin and treating atopic dermatitis. J Am Acad Dermatol. 2002; 46(2): 228–41. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5674", "date": "05 Aug 2014", "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\nIn the abstract and introduction sections the authors stated that solitary mastocytoma accounts for 10-15% of cutaneous mastocytosis whereas in the introduction section it was 10-20%. The lesion was present since birth, but parents sought medical advice after 18 months. Was there any particular reason for this delay? In particular, was any other previous treatment(s) prescribed for the child? The magnification of Fig, 2 was mentioned as 40X but it seems that Fig. 2A and 2B have different magnifications. Authors should also insert the magnification of Fig.3. As per any case report describing a treatment for a disorder known to be self-limited, it is uncertain whether the resolution is due to the applied drug or due to natural spontaneous subsidence.", "responses": [] }, { "id": "5676", "date": "19 Aug 2014", "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\nI thank the authors for this interesting article.I believe it is worthy of indexation because there is surprisingly little in the literature about this alternative treatment. I might make a few small suggestions for consideration:Is solitary mastocytoma really \"rare\"? I see an awful lot of them for such a designation even taking into consideration referral bias. Uncommon might be a better descriptor. Could the line \"Diagnosis is by biopsy...\" be modified? I do not think these absolutely must be biopsied and in fact almost never do. If there is a + Dariers sign and a strong clinical suspicion this presentation is specific enough that I do not think biopsy is mandatory. As a pediatric dermatologist I do all I can to avoid biopsying when not absolutely necessary and I worry readers might take this line to imply diagnosis mandatory for diagnosis of mastocytoma. It is not. Can we still call tacrolimus and pimecrolimus \"new\" given they have been available almost 15 years now? I think in fairness the authors must mention the boxed warning about this class of medications somewhere. If the authors cite the concerns for topical and systemic side effects of topical steroids as they do I think they must balance this by mentioning the biggest barrier to using these agents - the black box. I might have included two references the authors omitted:Correia O, Duarte AF, Quirino P et al.: Cutaneous mastocytosis: Two pediatric cases treated with topical pimecrolimus. Dermatology Online Journal. 2010;16(5): 8This describes two cases similar to the authors' treated successfully with pimecrolimus Avshalumov K, Pichardo R, Jorizzo JL et al.: Bullous mastocytosis: report of a patient and a brief review of the literature. Am J Dermatopath. 2008; 30(5):455-7Cutaneous mastocytosis - albeit not mastocytoma - treated with a number of things including topical tacrolimus.", "responses": [] }, { "id": "5872", "date": "19 Aug 2014", "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 described an 18-mo old girl with a solitary mastocytoma, which was successfully treated with tacrolimus. This might suggest that the mastocytoma requires therapy. This is not the case. Mastocytomas are self-limited and usually don´t need therapy. The most important management is the avoidance of known trigger factors. It is not clear whether the reduction is due to the self-limiting nature of the tumor or to the therapy. Pathogenetically and from the mechanism of action of tacrolimus, which prevents mast cell degranulation, an improvement of mastocytoma can be expected. However it should be made clear that the indication for treatment of a solitary mastozytoma should be made very cautiously.", "responses": [] } ]	1	https://f1000research.com/articles/3-181
https://f1000research.com/articles/3-178/v1	30 Jul 14	{ "type": "Research Article", "title": "Immunoprecipitation and mass spectrometry identify non-cell autonomous Otx2 homeoprotein in the granular and supragranular layers of mouse visual cortex", "authors": [ "Namsuk Kim", "Dario Acampora", "Florent Dingli", "Damarys Loew", "Antonio Simeone", "Alain Prochiantz", "Ariel A. Di Nardo", "Namsuk Kim", "Dario Acampora", "Florent Dingli", "Damarys Loew", "Antonio Simeone" ], "abstract": "Plasticity in the visual cerebral cortex is regulated by the internalization of Otx2 homeoprotein into parvalbumin neurons in cortical layers II/III and IV. However the Otx2 locus is not active in these neurons and the protein is imported from external sources, including the choroid plexus. Because Otx1 and Otx2 may have redundant functions, we wanted to verify if part of the staining in parvalbumin neurons corresponds to Otx1 transported from cortical layer V neurons. It is demonstrated here that Otx staining in layer IV cells is maintained in Otx1-null mice. The immunoprecipitation of extracts from finely dissected granular and supragranular cortex (layers I-IV) gave immunoblots with a band corresponding to Otx2 and not Otx1. Moreover, high-resolution mass spectrometry analysis after immunoprecipitation identifies two peptides within the Otx2 homeodomain. One of these peptides is specific for Otx2 and is not found in Otx1. These results unambiguously establish that the staining in parvalbumin neurons revealed with the anti-Otx2 antibodies used in our previous studies identifies non-cell autonomous Otx2.", "keywords": [ "Neural circuits generated during embryonic development are remodeled by environmental inputs during periods of heightened plasticity in postnatal development1. These critical periods are limited to specific windows of time that are different for each sensory system. In the visual system", "the primary visual cortex is subjected to a critical period for ocular dominance plasticity during which connections from a weak or absent eye can be permanently overtaken by the strong eye2. The integrated action between inhibitory and excitatory circuits determines critical period onset", "with a major role being played by the maturation of fast-spiking parvalbumin (FSPV) interneurons3. We have shown that Otx2 homeoprotein helps determine critical period timing by signaling FSPV cells in mice4. Conditional knock-down in heterozygous floxed mice just prior to normal critical period timing is sufficient to delay onset", "while cortical infusion of recombinant Otx2 protein accelerates both onset and closure4." ], "content": "Introduction\n\nNeural circuits generated during embryonic development are remodeled by environmental inputs during periods of heightened plasticity in postnatal development1. These critical periods are limited to specific windows of time that are different for each sensory system. In the visual system, the primary visual cortex is subjected to a critical period for ocular dominance plasticity during which connections from a weak or absent eye can be permanently overtaken by the strong eye2. The integrated action between inhibitory and excitatory circuits determines critical period onset, with a major role being played by the maturation of fast-spiking parvalbumin (FSPV) interneurons3. We have shown that Otx2 homeoprotein helps determine critical period timing by signaling FSPV cells in mice4. Conditional knock-down in heterozygous floxed mice just prior to normal critical period timing is sufficient to delay onset, while cortical infusion of recombinant Otx2 protein accelerates both onset and closure4.\n\nRemarkably, Otx2 protein in the cortex is non-cell autonomous. The Otx2 locus is silent, as shown by PCR, in situ hybridization and Otx2+/GFP mice, while Otx2 protein is detectable by immunohistochemistry and immunoblot4,5. We have shown that cortical infusion of recombinant Otx2 protein results in specific uptake by FSPV cells, while injection in the retina results in its transport along the visual pathway and into these same cells4. Blocking extracellular Otx2 through infusion of antibodies or specific peptides reduces uptake of endogenous Otx2 in FSPV cells4,6. Furthermore, we recently showed that the choroid plexus expresses Otx2 and secretes it into the cerebrospinal fluid5. Conditional knockdown of Otx2 expression in the choroid plexus results in reduced cortical levels of Otx2 protein. Needless to say, all of these approaches that alter Otx2 protein levels in the visual cortex have resulted in changes in cortical plasticity timing7.\n\nAn outstanding question is whether cortical Otx1 homeoprotein also plays a role in the critical period. Indeed, Otx1 is expressed in the cerebral cortex during development and continues to be expressed by layer V neurons in the adult8. It is thus possible that it is secreted by layer V cells and transferred into above granular and supragranular layers where Otx2 protein is detected. Unfortunately, most antibodies (commercial and academic) for Otx1 and Otx2 are pan-Otx thereby ruling out immunohistochemical approaches for conclusive evidence. Since genetic manipulation of the Otx2 locus has resulted in reduced protein in layer IV visual cortex7, we sought confirmation whether Otx1 is also present in these layers in the adult by using Otx1 knockout mice and proteomic approaches.\n\n\nMethods\n\nAll experiments were conducted in accordance with European Union Council Directives (86/609/EEC) and conform to Directive 2010/63/EU of the European Parliament. This study falls under project #00704.02 authorized by the French Ministry of Research. Adult male C57Bl/6J mice (Janvier) aged 2 to 3 months were used throughout this study and euthanized by cervical dislocation. Otx1 knockout mice, in which Otx1 is replaced with lacZ gene, have been previously described14. Adult (~3 months) male Otx1LacZ/LacZ mice were fixed by intracardiac perfusion.\n\nTotal RNA from frozen tissue samples were extracted by using RNeasy Mini kits (Qiagen) and were reverse transcribed (~400 ng) with Superscript II and oligo-dT primers (Invitrogen). For real-time PCR, samples were analyzed with a LightCycler Instrument (Roche) and SYBR green (Sigma) detection with the following primers:\n\nOtx2-forward ATTCACGTTTCATGACCAACAG;\n\nOtx2-reverse ATTGACTCCGTATGAGCGGTAT;\n\nOtx1-forward GAACCTTCCTTCTCCGAAATCT;\n\nOtx1-reverse GATCTTCACATCGGACAAATCA;\n\nGAPDH-forward TGACGTGCCGCCTGGAGAAAC;\n\nGAPDH-reverse CCGGCATCGAAGGTGGAAGAG.\n\nThe experiment was performed in duplicate. For calculating fold expression, gene-to-GAPDH ratios were determined by using the 2-∆∆Ct method referenced to Otx1 expression in the lateral geniculate nucleus (LGN).\n\nDissected tissue was lysed in 100 mM Tris-HCl, pH 7.4, 10 mM EDTA, 150 mM NaCl, 1% Triton X-100, and protease inhibitor (Roche) then triturated (22G and 26G needles). Lysates were centrifuged at 16,000 × g for 10 min at 4°C and supernatants were incubated either directly with antibodies or with antibody-coupled Dynabeads (Life Technologies) at 4°C for 16 h. For samples with uncoupled antibodies, Protein A Dynabeads (Life Technologies) were added and incubated at 4°C for 1 h. Proteins were eluted with 2X SDS sample buffer after five washes with lysis buffer. Antibodies were rabbit polycolonal IgG (Abcam ab27478) and anti-Otx2 (rabbit polyclonal, Abcam ab21990) and were coupled to Dynabeads according to manufacturer instructions (Life Technologies). For immunoblots, samples were separated on NuPAGE 4%–12% Bis-Tris precast gels (Invitrogen) and transferred onto PVDF membrane. Membranes were incubated overnight at 4°C with anti-Otx2 (rabbit polyclonal, 1/1,000, Abcam ab21990) and then with HRP-coupled anti-rabbit IgG (1/2,000, GE Healthcare NA934) 1 h at RT.\n\nFor immunostaining, 50 µm floating sections were incubated with anti-Otx2 (rat polyclonal, 1/200, in-house) and biotinylated-WFA (1/100, Sigma L1516) in TBS, 1% Triton-X, 0.2% Tween-20, 10% fetal calf serum, overnight at 4°C. Sections were extensively washed at RT, incubated 2 h at RT with anti-rat Alexa Fluor-488 (Molecular Probes A21208, 1/2,000) and streptavidin Alexa Fluor-546 (Molecular Probes S11225, 1/2,000), washed again and finally mounted in Fluoromount medium (SouthernBiotech). Images were acquired with an Eclipse 90i microscope (Nikon).\n\nAfter immunoprecipitation, proteins were separated on SDS–PAGE gels (Invitrogen) and stained with colloidal blue staining (LabSafe GEL BlueTM GBiosciences). Gel slices were excised and proteins were reduced with 10 mM DTT prior to alkylation with 55 mM iodoacetamide. After washing and shrinking the gel pieces with 100% MeCN, in-gel digestion was performed using trypsin (Promega) overnight in 25 mM NH4HCO3 at 30°C.\n\nPeptides were extracted and analyzed by nano-LC-MS/MS using an Ultimate 3000 system (Dionex S.A.) coupled to an Orbitrap Fusion mass spectrometer (Q-OT-qIT, Thermo Fisher Scientific). Samples were loaded on a C18 precolumn (300 µm inner diameter × 5 mm; Dionex) at 20 µl/min in 5% MeCN, 0.1% TFA. After a desalting for 3 min, the precolumn was switched on the C18 column (75 μm i.d. × 50 cm, packed with C18 PepMap™, 3 μm, 100 Å; LC Packings) equilibrated in solvent A (2% MeCN, 0.1% HCO2H). Bound peptides were eluted using a 150 min linear gradient (from 5 to 30% (v/v)) of solvent B (80% MeCN, 0.085% HCO2H) at a 150 nl/min flow rate and an oven temperature of 40°C. We acquired Survey MS scans in the Orbitrap on the 400–1500 m/z range with the resolution set to a value of 240,000 and a 4 × 105 ion count target. Each scan was recalibrated in real time by co-injecting an internal standard from ambient air into the C-trap. Tandem MS was performed by isolation at 1.6 Th with the quadrupole, HCD fragmentation with normalized collision energy of 35, and rapid scan MS analysis in the ion trap. The MS2 ion count target was set to 104 and the max injection time was 200 ms. Only those precursors with charge state 2–7 were sampled for MS2. The dynamic exclusion duration was set to 60 s with a 10 ppm tolerance around the selected precursor and its isotopes. The instrument was run in top speed mode with 3 s cycles.\n\nData were acquired using the Xcalibur software (v 3.0.63) and the resulting spectra were interrogated by the MascotTM Software through Proteome Discoverer (v 1.4.0.288, Thermo Scientific) with the SwissProt Mus musculus database (20140402, 16,671 sequences). We set carbamidomethyle cysteine, oxidation of methionine and N-terminal acetylation as variable modifications. We set specificity of trypsin digestion and allowed 2 missed cleavage sites and we set the mass tolerances in MS and MS/MS to 2 ppm and 0.5 Da, respectively. The resulting Mascot files were further processed by using myProMS (v 3.0)15 and the estimated false discovery rate (FDR) by automatically filtering the Mascot score of all peptide identifications was less than 0.5%.\n\n\nResults\n\nWe compared the expression of the Otx1 and Otx2 loci of various brain regions by performing quantitative PCR (Figure 1A). While both mRNA were detected in thalamic and cerebellar structures, only Otx1 mRNA was found in the visual cortex. This result confirms the previously reported absence of GFP expression in the visual cortex of Otx2+/GFP knockout mice and lack of signal in the visual cortex of wild type mice after Otx2 in situ hybridization, even though Otx2 antibodies label FSPV cells4. These cells are enwrapped by a dense extracellular matrix called perineuronal nets (PNNs) when localized to layer IV of the cortex (Figure 1B). Otx1 locus is active in layer V8. In immunohistochemical analysis of Otx1 knockout mouse, almost all Otx signal is lost in layer V while signal from Otx2 protein continues in PNN-labeled cells (Figure 1B). However, we cannot preclude that some Otx1 is also present in layer IV cells in wild type mice.\n\n(A) Analysis of Otx1 and Otx2 expression by quantitative RT-PCR on extracts from lateral geniculate nucleus (LGN), visual cortex (V Cx), superior colliculus (S Col) and cerebellum (Cb). The fold-difference in expression is calculated relative to Otx1 in LGN. The Otx2 locus is silent in visual cortex. (B) Non-cell autonomous Otx2 is found in visual cortex. Immunostaining in wild type mice reveals Otx1/2 cells in layers IV and V of visual cortex, including cells with perineuronal nets (stained by WFA lectin) enriched in layer IV. Staining for Otx2 persists in Otx1 null mice (Otx1 KO). Scale bar, 50 µm.\n\nIn order to analyze Otx homeoprotein distribution in the visual cortex, we turned to immunoprecipitation (IP) experiments. We first performed IP on whole visual cortex extracts, which showed both Otx1 and Otx2 protein (Figure 2A). This result is confirmed by IP of choroid plexus, which strongly expresses Otx2 but only very weakly expresses Otx1. To analyze granular and supragranular content, we dissected and extracted the superior layers of posterior adult mouse cortex (Figure 2B) and performed IP by using cross-linked magnetic beads. Immunoblot analysis detected only Otx2 but not Otx1 protein (Figure 2C). This result was confirmed by mass spectrometry analysis on these extracts, which identified 2 Otx2 peptides (6.6% coverage with 100% specificity, Figure 2D). While this number of peptides is low, it was expected given that Otx2 is predicted to be poorly ionizable. Furthermore, identification has an error of 1 protein in 20,000 (19,999 are true) with a FDR of less than 0.5% for 2 peptides in our database of 16,671 sequences, thus the chance of misinterpretation is very low. These peptides are 100% specific for Otx2 and they have the same ions as the reference spectrum from a sample of purified Otx2 protein. While the second peptide is also specific for Otx1 and Crx homeoproteins, the first peptide is unique for Otx2 (Figure 2D). These results confirm that the homeoprotein localized in granular and supragranular FSPV cells is indeed Otx2 and not Otx1.\n\n(A) Immunoblots for Otx1/2 of immunoprecipitation (IP) experiments on extracts from visual cortex and choroid plexus. (B) Diagram of finely dissected region for extracts containing granular (IV) and supragranular (I-III) layers of visual cortex. (C) Immunoblot for Otx1/2 of samples from IP using cross-linked magnetic beads and finely dissected extracts. Choroid plexus (Ch Pl) extract was used to control for Otx2 migration velocity. (D) The peptides (red, bold) matching Otx2 protein identified by high-resolution mass spectrometry. Only the homeodomain sequence of Otx2 (amino acids 38–97) is shown. The amino acid differing in Otx1 is highlighted in green, while amino acids that differ in Crx are highlighted in green and in yellow.\n\n\nDiscussion\n\nThe non-cell autonomous activity of homeoprotein transcription factors is now well established. There are clear phenotypes with recently developed in vivo single-chain secreted antibodies that neutralize extracellular homeoproteins yet leave intact cell autonomous activities9–11. Non-autonomy can also be demonstrated by comparing mRNA and protein expression. Indeed, the absence of mRNA in presence of the protein argues in favor of non-cell autonomy. However, when the receiving territory is a short distance from the producing territory, one could invoke the possibility of cell migration or mRNA instability to bring into question the reality of homeoprotein transfer.\n\nIn the visual system, Otx2 protein is found in the visual cortex far from two potential sources of Otx2 (where the Otx2 locus is active), namely the eye and the choroid plexus5,12. Indeed, the Otx2 locus is not active in the adult cerebral cortex as verified by using the Otx2+/GFP mouse, quantitative RT-PCR and in situ hybridization4. In addition, conditional Otx2 ablation in the choroid plexus reduces its content in FSPV cells, further supporting non-cell autonomy5.\n\nHowever, since most Otx1 and Otx2 antibodies are pan-Otx antibodies, it was still conceivable that some of the protein seen in FSPV cells by immunohistochemistry could correspond to Otx1 expressed in layer V of the cerebral cortex and transferred into PV cells. The present study shows that the staining in layer IV is maintained in the Otx1 knockout mouse and that IP experiments of layers I-IV give immunoblot bands with expected Otx2 size and can be used to identified Otx2 by mass spectrometry. These results confirm that FSPV cells in granular and supragranular layers of the cerebral cortex only contain non-cell autonomous Otx2 and do not contain Otx1.\n\nIt may seem surprising that Otx1 expressed in layer V is not secreted and internalized by FSPV cells. Indeed the protein presents a homeodomain nearly identical to that of Otx2 and thus contains the two sequences necessary for internalization and secretion (for review see7). However, previous studies have demonstrated that homeoproteins are transported from the basolateral to the apical side of polarized cells and thus into the axon7,13. Given their polarity and orientation, the pyramidal cells of layer V that express Otx1 are thus very unlikely to release it at the level of FSPV cells. In contrast, the choroid plexus epithelial cells present their apical surface toward the ventricles allowing Otx2 secretion into the cerebral spinal fluid. In conclusion, this study demonstrates that Otx2 is the only non-cell autonomous Otx family protein in the granular and supragranular FSPV cells.\n\n\nData availability\n\nF1000Research: Dataset 1. Quantitative PCR and mass spectrometry data of Otx2 homeoprotein in the mouse visual cortex, 10.5256/f1000research.4869.d3338416", "appendix": "Author contributions\n\n\n\nNK and AAD carried out experimental work\n\nDA and AS made Otx1 knockout mice\n\nDF carried out the mass spectroscopy experimental work\n\nLD supervised mass spectroscopy and proteomic data analysis\n\nAP and AAD conceived the ideas of the study, designed protocols, and drafted the manuscript\n\nAll authors read, critically revised, and approved the final manuscript\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Région Ile-de-France, the FRM, GRL Program n°2009-00424, ANR grant BRAINEVER n° 11-BLAN-069467, and ERC Advanced Grant HOMEOSIGN n°339379.\n\n\nReferences\n\nHensch TK: Critical period plasticity in local cortical circuits. Nat Rev Neurosci. 2005; 6(11): 877–88. PubMed Abstract \| Publisher Full Text\n\nMorishita H, Hensch TK: Critical period revisited: impact on vision. Curr Opin Neurobiol. 2008; 18(1): 101–7. PubMed Abstract \| Publisher Full Text\n\nHensch TK, Fagiolini M: Excitatory-inhibitory balance and critical period plasticity in developing visual cortex. Prog Brain Res. 2005; 147: 115–24. PubMed Abstract \| Publisher Full Text\n\nSugiyama S, Di Nardo AA, Aizawa S, et al.: Experience-dependent transfer of Otx2 homeoprotein into the visual cortex activates postnatal plasticity. Cell. 2008; 134(3): 508–20. PubMed Abstract \| Publisher Full Text\n\nSpatazza J, Lee HHC, Di Nardo AA, et al.: Choroid-plexus-derived Otx2 homeoprotein constrains adult cortical plasticity. Cell Rep. 2013; 3(6): 1815–23. PubMed Abstract \| Publisher Full Text\n\nBeurdeley M, Spatazza J, Lee HHC, et al.: Otx2 binding to perineuronal nets persistently regulates plasticity in the mature visual cortex. J Neurosci. 2012; 32(27): 9429–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpatazza J, Di Lullo E, Joliot A, et al.: Homeoprotein signaling in development, health, and disease: a shaking of dogmas offers challenges and promises from bench to bed. Pharmacol Rev. 2013; 65(1): 90–104. PubMed Abstract \| Publisher Full Text\n\nFrantz GD, Weimann JM, Levin ME, et al.: Otx1 and Otx2 define layers and regions in developing cerebral cortex and cerebellum. J Neurosci. 1994; 14(10): 5725–40. PubMed Abstract\n\nLesaffre B, Joliot A, Prochiantz A, et al.: Direct non-cell autonomous Pax6 activity regulates eye development in the zebrafish. Neural Dev. 2007; 2: 2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWizenmann A, Brunet I, Lam JSY, et al.: Extracellular Engrailed participates in the topographic guidance of retinal axons in vivo. Neuron. 2009; 64(3): 355–66. PubMed Abstract \| Publisher Full Text\n\nLayalle S, Volovitch M, Mugat B, et al.: Engrailed homeoprotein acts as a signaling molecule in the developing fly. Development. 2011; 138(11): 2315–23. PubMed Abstract \| Publisher Full Text\n\nSugiyama S, Prochiantz A, Hensch TK: From brain formation to plasticity: insights on Otx2 homeoprotein. Dev Growth Differ. 2009; 51(3): 369–77. PubMed Abstract \| Publisher Full Text\n\nDupont E, Prochiantz A, Joliot A: Identification of a signal peptide for unconventional secretion. J Biol Chem. 2007; 282(12): 8994–9000. PubMed Abstract \| Publisher Full Text\n\nAcampora D, Mazan S, Avantaggiato V, et al.: Epilepsy and brain abnormalities in mice lacking the Otx1 gene. Nat Genet. 1996; 14(2): 218–22. PubMed Abstract \| Publisher Full Text\n\nPoullet P, Carpentier S, Barillot E: myProMS, a web server for management and validation of mass spectrometry-based proteomic data. Proteomics. 2007; 7(15): 2553–6. PubMed Abstract \| Publisher Full Text\n\nKim N, Acampora D, Dingli F, et al.: Quantitative PCR and mass spectrometry data of Otx2 homeoprotein in the mouse visual cortex. F1000Research. 2014. Data Source" }	[ { "id": "5640", "date": "04 Aug 2014", "name": "Fred Gage", "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 address an issue that is unresolved in the literature. Is the Otx found in the cortex, and responsible in part of cortical plasticity, cortical derived Otx1 or choroid plexus derived Otx2? The problem exists because the antibodies used previously can not distinguish between the two. In the present study using Otx1 KO mice and a variety of measures, the authors prove Otx2 is the protein in the cortical cells and that this therefore is a cell non- autonomous phenomenon. The one question I have relates to figure 1B, which suggests there is greater Otx2 staining in the Otx1 KO mouse cortex. If this is a reliable observation the authors should comment. The low peptide coverage in the MASS SPEC could be a concern but the authors address that and in addition the other supporting data decrease the concern.This is a clear, straightforward manuscript that adds important and definitive information to the existing literature", "responses": [] }, { "id": "5636", "date": "08 Aug 2014", "name": "Paola Bovolenta", "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\nStudies, mostly from A. Prochiantz’s laboratory, have demonstrated that the transcription factor (TF) Otx2, produced by external sources such as the choroid plexus, is internalized into parvalbumin neurons of layers II/III and IV of the visual cortex, thereby controlling its plasticity. So far these studies had not ruled out the possibility that part of this effect could be mediated by Otx1, a closely related TF, which may act redundantly with Otx2. In this study, the authors combined the use of Otx1-/- mice with immunoprecipitation from layers I-IV extracts and high-resolution mass spectrometry analysis to demonstrate the sole presence of Otx2 peptides in these layers. The study thus demonstrates that Otx2 is the only non-cell autonomous Otx family member functional in fast-spiking parvalbumin cells of layers I-IV of the visual cortex. This is a clear and well-executed study that solves an existing concern.", "responses": [] } ]	1	https://f1000research.com/articles/3-178
https://f1000research.com/articles/3-177/v1	30 Jul 14	{ "type": "Software Tool Article", "title": "dendsort: modular leaf ordering methods for dendrogram representations in R", "authors": [ "Ryo Sakai", "Raf Winand", "Toni Verbeiren", "Andrew Vande Moere", "Jan Aerts", "Raf Winand", "Toni Verbeiren", "Andrew Vande Moere", "Jan Aerts" ], "abstract": "Dendrograms are graphical representations of binary tree structures resulting from agglomerative hierarchical clustering. In Life Science, a cluster heat map is a widely accepted visualization technique that utilizes the leaf order of a dendrogram to reorder the rows and columns of the data table. The derived linear order is more meaningful than a random order, because it groups similar items together. However, two consecutive items can be quite dissimilar despite proximity in the order. In addition, there are 2n-1 possible orderings given n input elements as the orientation of clusters at each merge can be flipped without affecting the hierarchical structure. We present two modular leaf ordering methods to encode both the monotonic order in which clusters are merged and the nested cluster relationships more faithfully in the resulting dendrogram structure. We compare dendrogram and cluster heat map visualizations created using our heuristics to the default heuristic in R and seriation-based leaf ordering methods. We find that our methods lead to a dendrogram structure with global patterns that are easier to interpret, more legible given a limited display space, and more insightful for some cases. The implementation of methods is available as an R package, named ”dendsort”, from the CRAN package repository. Further examples, documentations, and the source code are available at [https://bitbucket.org/biovizleuven/dendsort/].", "keywords": [ "Agglomerative hierarchical clustering (HC) is one of the classic and yet still very popular cluster analysis methods in data exploration1", "2. Its implementation is widely available and execution of the clustering requires only a few settings", "such as a choice of distance metric and linkage algorithm3. The clustering process begins with individual input elements as singleton clusters and successively merges a pair of most similar clusters until only one cluster remains. The dissimilarity", "or the distance", "between two clusters is defined by a distance metric and updated by a linkage algorithm. The output of HC is typically represented in a form of a binary tree", "called a dendrogram. In a dendrogram", "the similarity of two clusters is encoded in the height of the branch where two clusters merge. Two very similar elements are merged in the early stages of clustering", "thus the height of the branches between these elements is relatively small. The dissimilarity between two clusters increases with each successive merge", "resulting in a binary hierarchical structure with a monotonic property4. Therefore", "a dendrogram represents both cluster-subcluster relationships as well as the order in which the clusters were merged5." ], "content": "Introduction\n\nAgglomerative hierarchical clustering (HC) is one of the classic and yet still very popular cluster analysis methods in data exploration1,2. Its implementation is widely available and execution of the clustering requires only a few settings, such as a choice of distance metric and linkage algorithm3. The clustering process begins with individual input elements as singleton clusters and successively merges a pair of most similar clusters until only one cluster remains. The dissimilarity, or the distance, between two clusters is defined by a distance metric and updated by a linkage algorithm. The output of HC is typically represented in a form of a binary tree, called a dendrogram. In a dendrogram, the similarity of two clusters is encoded in the height of the branch where two clusters merge. Two very similar elements are merged in the early stages of clustering, thus the height of the branches between these elements is relatively small. The dissimilarity between two clusters increases with each successive merge, resulting in a binary hierarchical structure with a monotonic property4. Therefore, a dendrogram represents both cluster-subcluster relationships as well as the order in which the clusters were merged5.\n\nThere are two unique uses of a dendrogram in exploratory data analysis. First, clusters of input elements can be inferred from the subtree structures below a certain threshold by “cutting the tree”. It is an advantage of hierarchical clustering that this threshold value can be adjusted based on domain-specific knowledge to result in clusters of different sizes. Second, a linear order of observations (rows) or attributes (columns) of an associated matrix can be derived. This linear order of observations is typically used to reorder the columns or rows of the data matrix. Then, the matrix is visualized as cluster heat maps1, where dendrograms and heat map visualizations are coupled (Figure 1).\n\nThe linear order derived from a dendrogram is more meaningful than a random order, as it groups similar items together6,7. However, two consecutive items in this order are not necessarily similar, since these leaves could belong to different subtree structures, or simply be quite distant from each other. This is a common misinterpretation of a dendrogram: one may expect similarity between two input elements based on the proximity in the leaf order8,9. In addition, there are 2n−1 possible orderings given n input elements, because the orientation of clusters at each merge can be flipped without affecting the underlying hierarchical structure, thus rendering a unique optimization challenge.\n\nTo address the misinterpretation of dendrograms and the optimization problem, a number of methods have been proposed to rearrange the structure of a dendrogram. Gruvaeus and Wainer10 proposed a method (GW) to order leaves such that two singleton clusters at the edge of adjacent subtrees are most similar, given the constraint of the binary tree structure. Bar-Joseph et al.6 proposed a method, called the optimal leaf ordering (OLO), to maximize the sum of the similarity of any adjacent elements in the ordering. Similarly, Chae and Chen11 proposed a method for ordering by minimizing the bilateral symmetric distance between two adjacent clusters. All these methods aim to homogenize the linear order in one way or another and are evaluated in terms of either a loss function, such as the Hamiltonian path length, or a merit function, such as the number of anti-Robinson events12.\n\nEven though these seriation-based leaf ordering methods exploit the binary tree structure to reduce the number of permissible permutations, these methods have short-comings. First, they homogenize and optimize the distance between items in the linear order, and this still encourages the common misinterpretation of dendrograms, reading a dendrogram horizontally. Second, the dendrogram structure is only a means to reduce the number of permissible permutations, and the graphical representation of the resulting dendrogram obscures the intrinsic properties of the hierarchical clustering result, such as the cluster-subcluster relationship and the order in which clusters are merged.\n\nIn the biological domain, Eisen et al.13 have introduced and established a cluster analysis method for high throughput gene expression data using cluster heat maps. The method includes a leaf orderings by weighting genes based on genome coordinates or the average expression level. The resulting linear order is more meaningful in terms of biology, but the method requires prior knowledge or additional information for the weighting.\n\nIn this paper, we present leaf ordering heuristics, named modular leaf ordering (MOLO), to address the aforementioned shortcomings by constructing a dendrogram that reflects a) the monotonic order in which clusters are merged and b) the nested cluster relationships. We compare dendrogram and cluster heat map visualizations created using our heuristics to the default heuristic in R and seriation-based leaf ordering methods. The implementation is available as an R package, named \"dendsort\", from the CRAN package repository. The R script for generating figures in this paper is available as a supplementary material. Further examples, documentations, and the source code are available at [https://bitbucket.org/biovizleuven/dendsort/].\n\n\nMethods\n\nAgglomerative hierarchical clustering (HC) starts with individual observations as singleton clusters and merges clusters iteratively until all clusters belong to one big cluster. In each iteration, the two most similar clusters are identified by a distance measure and a linkage algorithm of choice. The details of the algorithm and the properties of distance measures and linkage algorithms are described in4,5,14.\n\nThe default hierarchical clustering method in R combines three types of merges: a merge between two singleton clusters, a merge between a singleton cluster and a cluster with more than one member, and a merge between clusters with multiple members. The heuristics for determining the orientation of merging elements essentially determine the structure of the resulting dendrogram.\n\nUsing a simple two-dimensional data set as shown in Figure 2A, we demonstrate the default heuristics used in the hierarchical clustering method in R. A dendrogram is constructed as follows: When a leaf (singleton cluster) merges with another leaf, the orientation of clusters is determined by the order of observations in the input data matrix, as seen in branch “a”, “b”, “c” and “f” in Figure 2B. When a leaf merges with a cluster with more than one member (subtree), the leaf is always placed on the left side of the branch, as shown in branch \"d\" and \"g\". When two subtree merges, the subtree with the smaller distance in the previous merge is placed on the left, as seen in branch “e”, “h”, and “i”. Each branch is labeled alphabetically in the order of merges within the clustering process.\n\n(A) A scatterplot of the ten input elements. The number of each element also represents the order in the input matrix. (B) A dendrogram drawn using the default heuristics in R. The branches in the dendrogram are labeled from “a” to “i” in the order in which clusters are merged. (C) A dendrogram reordered using MOLO with the smallest distance. The global structures in a shape of the right triangle are highlighted.\n\nIn contrast to the default heuristics, our heuristics are characterized by 2 key differences: first, a leaf is placed on the right side when it merges with a subtree; second, when two subtrees merge, the subtree with the smallest distance among all of preceding merges is placed on the left (Figure 2C). The first rule avoids a branch of a singleton cluster hanging over the preceding nested clusters and allows the tree to grow from left to right in the order of merges. The second rule ensures that the tightest cluster is placed leftmost within the subtree structure. Consequently, our heuristics result in each subtree or sub-cluster structure in a right triangular shape, as shown in Figure 2C. This feature increases the contrast between the items at the edge of adjacent subtree structures, thus modularizing each subtree structure.\n\nThe MOLO method takes the result of the default hierarchical clustering method, and reevaluates the orientation of the clusters at each branch recursively. The pseudocode of this algorithm is shown in Figure 3. In addition to the algorithm based on the smallest distance, we also implemented a variant in which the average distances of all preceding merges are compared, and discussed further in the third case study. The data in Figure 2 consist of only 10 observations and it is merely intended to explain the difference in heuristics. Following case studies demonstrate applications of the MOLO algorithm with larger datasets, and compare visualizations created using our heuristics and other existing leaf ordering methods.\n\n\nResults\n\nOne of the key tasks in applying hierarchical clustering is to choose an appropriate distance metric and a linkage algorithm14. A choice of distance metric, such as Euclidean distance and correlation-based distance, defines a measure of similarity between two elements. Clustering algorithms, such as complete, average, and single linkage, are variations of the cluster proximity definition5. The choice of distance measures and linkage algorithms influences the clustering results. It is therefore recommended to try different HC settings in exploratory data analysis, especially when the underlying data structure is unknown.\n\nAs Hastie et al.4 point out, dendrogram structures can vary greatly depending on the choice of linkage algorithms. In Figure 4, dendrograms of different linkage algorithms for the same simulated data set are compared. The appearance of the dendrogram structure is quite different and it is difficult to compare similarities in the nested cluster structure. In contrast, when the MOLO method is applied, we find the reordered dendrograms easier to study the nested structure and to compare between one another (Figure 5), because the linear leaf order in these dendrograms reflect the order in which clusters are merged. For instance, the element 32 and 34 form the tightest cluster, and they are easy to identify because they are always placed leftmost. Also, upon closer examination of the reordered dendrogram structures, we find that the reordered dendrograms reflect the underlying difference in algorithms more closely. For example, the average linkage is an intermediate approach between the single and complete linkage algorithms to define cluster proximity5. Although the MOLO method does not change the clustering result itself, this case study demonstrates how it can improve, or at least bring a new perspective, to interpret dendrogram structures.\n\nThe element 32 and 34 are highlighted.\n\nThe element 32 and 34 are highlighted.\n\nThe second case study extends the demonstration of seriation-based leaf ordering methods by Buchta et al.15 using the Fisher’s Iris data set. The Fisher’s Iris data set is available from the R’s dataset package16. This Iris data set represents 3 species of iris with 50 observations for each species. Each observation contains measurements of 4 attributes: the sepal length and width, and the petal length and width. In this case we performed hierarchical clustering on the distance matrix of Euclidean distances, using the complete linkage algorithm. In Figure 6, adjacency matrices are visualized as cluster heat maps to compare results of the default hierarchical clustering (HC), the Gruvaeus and Wainer’s method (GW)10, the optimal leaf ordering (OLO)6, and the MOLO method (MOLO). These matrices are diagonally symmetric and rows and columns are reordered based on the leaf order of dendrograms. The species for each observation is color coded and shown between the dendrogram and the heat map visualization. Implementations of the GW and OLO methods are available in the seriation R package15.\n\nThe default hierarchical clustering (HC), the Gruvaeus and Wainer’s method (GW), the optimal leaf ordering (OLO), and the MOLO method are applied to the Fisher’s Iris data set.\n\nDespite the fact that each representation shares the same underlying hierarchical clustering output, the visual impressions of heat maps are different depending on the choice of leaf ordering methods. For example, the results of the HC, GW, and OLO methods suggest two predominant clusters, as indicated by dark square blocks along the diagonal axis. On the other hand, the result of the MOLO method suggests three clusters. The MOLO heuristics place the most similar items on the left ends of each subtree structure and subsequently merged clusters are placed on its right. As a result, the MOLO method reorders the dendrogram structure to reflect the modularity of the cluster-subcluster structure. With the information of species for each observation, it becomes clear that there are three species and a half of versicolor samples are clustered together with virginica.\n\nAdditionally, we find the cluster edges in the heat map visualization of the MOLO method are more prominent than those of other leaf ordering methods. One explanation for the enhanced edges is the increased contrast between subtree structures, whereas the GW and OLO methods aim to reduce the edge contrast between sub-tree structures, resulting in more fuzzy boundaries. This effect can be seen at the borders between versicolor and virginica species in heat map visualizations. The second explanation is that the monotonic linear order results in an optical illusion, called Mach band effect, at the edge of subtree structures. The Mach band effect explains how edges in different shades of gray have exaggerated contrast when in contact17. This enhanced edge-detection works to our advantage in identifying clusters, especially because our visual systems to decode quantitative or continuous data from different shades of colors is limited18.\n\nAs also pointed out in previous studies6, the GW and OLO methods result in a global structure where highly similar items appear in the middle, while marginally related items are on the edge of the subtree structure. This tendency is most apparent in the setosa samples. On the other hand, the MOLO method results in a right triangular global shape where the similarity of clusters increases from left to right, unidirectionally, for each subtree structure. This global property enhances the contrast at the borders of clusters and reveals the third cluster in the heatmap visualizations.\n\nThe third case study involves a multivariate table obtained from the integrated pathway analysis of gastric cancer from the Cancer Genome Atlas (TCGA) study19. In this data set, each column represents a pathway consisting of a set of genes and each row represents a cohort of samples based on specific clinical or genetic features. For each pair of a pathway and a feature, a continuous value of between 1 and -1 is assigned to score positive or negative association, respectively. The goal of this cluster analysis is to explore patterns in the data set and examine clusters to characterize the link between the gene expression levels and clinical features and to identify subtypes of the cancer among the cohort of samples.\n\nThese matrices are typically visualized as cluster heat maps (Figure 1). By applying hierarchical clustering on the rows and columns independently, the rows and columns are reordered to place similar items close to each other. In this example, the distance measure is based on the Pearson’s Correlation coefficient and the complete linkage algorithm is used for hierarchical clustering.\n\nSimilarly to previous examples, the application of the MOLO method results in a global right triangular shape for each subtree, encoding the monotonicity of the hierarchical clustering process (Figure 7). However, upon a closer examination, we find that the first subtree of the rows does not form a right triangular shape. This first cluster is a very loose cluster having relatively long branches, except for the very first two rows which have the shortest distance. The characteristic of a loose cluster is also reflected in the heat map visualization, where there are no strong patterns of clustering, except for the first two rows. In order to prioritize tighter clusters with a smaller average distance, we implemented a variation of the modular leaf ordering method based on the average distance of the preceding merges (MOLO_AVG). The effects of leaf ordering methods on dendrogram structures for the rows are compared in Figure 8. With the MOLO_AVG method, the tight clusters with lower average distances are placed leftmost.\n\nThe rows from the example data sets are shown.\n\nThe cluster heat map generated with the MOLO_AVG method is shown in Figure 9. The choice of either the smallest or average distance does not influence the structure within subtrees, however the order of the subtree structures changes. Although the difference may be subtle, we find that the modularity of clusters becomes more distinctive with the MOLO_AVG method. The resulting visualization also provides new insights into relationships between clusters. For instance, the inverse relationship between sets of rows and columns becomes more apparent in Figure 9 than the original figure (Figure 1).\n\nThe rows and columns with an inverse relationship are highlighted in the dendrograms.\n\nOne way to evaluate the efficiency of a graphical representation is to compare the proportion of ink used to represent the data, a concept known as the data-ink ratio20. Since each dendrogram shares the same underlying hierarchical clustering output, the total length of lines required to draw a dendrogram can be directly compared to evaluate the conciseness of dendrogram representations. We calculated the total length of lines used to draw dendrograms in Figure 9, the results of which are shown in Table 1.\n\nThe MOLO_AVG method results in the highest reduction in the data-ink ratio, while the GW method results in an increase in the data-ink ratio. Since the total number of vertical lines in each dendrogram is the same, the difference in the total length is due to the horizontal lines. A factor contributing to the reduction of horizontal lines is the heuristic of placing the singleton cluster on the right side of the branch. This heuristic avoids the placement of a singleton cluster on the left side, spreading over the nested tree structure.\n\nAs the data size increases, the number of rows or columns in the data matrix increases while the display space for the figure may be limited. As a result, a dendrogram representation may become denser with more leaves, making the details of hierarchical structure harder to read. Figure 10 shows the same dendrograms as in Figure 8, but in a more limited display space. Because the MOLO methods results in a global pattern of right triangular shapes, it supports the viewer to identify tight and loose clusters even when the vertical lines of branches are so dense that they are in contact with adjacent branches. Similarly, because of this right triangular shape, each subtree structure is still distinguishable. Therefore, the MOLO methods aid the readability of dendrogram structures, even when the display size is limited.\n\nThe rows from the example data sets are shown.\n\nIn summary, this case study demonstrates how the MOLO methods support tasks in exploratory data analysis and improve readability of the dendrogram representations by reducing visual clutter. The dendrogram structure after the MOLO methods results in right triangular shapes for each subtree structure, and the order of leaves in each subtree reflects the order in which clusters are merged. In common with the case study of the Iris data set, the MOLO methods aid cluster identification in cluster heat maps.\n\n\nDiscussion\n\nIn this paper, we introduce two modular leaf ordering methods and demonstrate how leaf ordering of dendrograms can influence the interpretation of cluster heat map visualizations. While seriation-based leaf ordering methods focus on homogenizing the linear order of leaves, our heuristics focus on improving the graphical representation of dendrograms to reflect the intrinsic properties of the hierarchical clustering process, such as the monotonic increase of distances in successive merges. As a result, each subtree structure has a global right triangular shape. This modular property is also reflected in the linear order of leaves, thus influencing the visual impression of clusters in heat map visualizations.\n\nAlthough the leaf ordering methods affect the dendrogram representation and the linear order of leaves, it does not change the underlying hierarchical structure. In other words, the quality of the clustering results ultimately depends on the quality of the input data and the choice of appropriate distance metric and linkage algorithm. Given no prior knowledge of underlying patterns in data sets, it is recommended to try different normalization techniques in preprocessing and different distance measures and linkage algorithms to allow different aspects of the data to be explored14.\n\n\nConclusions\n\nThrough case studies, we demonstrate the effects of our leaf ordering methods on the interpretation of the clustering result, as well as the reduction in visual clutter as measured by the data-ink ratio. With cluster heat map techniques being very popular in life sciences, we advocate our methods to be considered both for exploratory data analysis and for publication of figures.\n\n\nSoftware availability\n\nhttp://cran.r-project.org/web/packages/dendsort/index.html\n\nhttps://bitbucket.org/biovizleuven/dendsort/\n\nhttps://bitbucket.org/F1000Research/dendsortarchive\n\nhttp://dx.doi.org/10.5281/zenodo.1098021\n\nGPL-2 \| GPL-3", "appendix": "Author contributions\n\n\n\nAll authors contributed to the design and organization of the paper and its writing and editing. RS initiated the project. TV and RS implemented the methods in R. RW and AVdM were involved in discussions for the development and JA supervised the project.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was performed under the umbrella of the KU Leuven Data Visualization Lab (www.datavislab.org) and supported through funding from the KU Leuven Research Council CoE PFV/10/016 SymBioSys (RS), the Academische Stichting Leuven vzw (RS), the IWT O&O ExaScience Life Pharma (TV), and iMinds ICON b-SLIM (RW).\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 Sheila Reynolds and Vésteinn Þórsson from the Institute for Systems Biology for sharing the sample data set for the third case study.\n\n\nReferences\n\nWilkinson L, Friendly M: The History of the Cluster Heat Map. Am Stat. 2009; 63(2): 179–184. Publisher Full Text\n\nGehlenborg N, O’Donoghue SI, Baliga NS, et al.: Visualization of omics data for systems biology. Nat Methods. 2010; 7(3 Suppl): S56–68. PubMed Abstract \| Publisher Full Text\n\nde Souto MC, Costa IG, de Araujo DS, et al.: Clustering cancer gene expression data: a comparative study. BMC Bioinformatics. 2008; 9: 497. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHastie T, Tibshirani R, Friedman J: The Elements of Statistical Learning. Springer Series Statistics. 2009. Publisher Full Text\n\nTan P, Kumar V, Steinbach M: Introduction to data mining. Boston: Pearson Addison Wesley, 1st ed edition. 2005. Reference Source\n\nBar-Joseph Z, Gifford DK, Jaakkola TS: Fast optimal leaf ordering for hierarchical clustering. Bioinformatics. 2001; 17(Suppl 1): S22–9. PubMed Abstract \| Publisher Full Text\n\nGehlenborg N, Wong B: Points of view: Heat maps. Nat Methods. 2012; 9(3): 213. Publisher Full Text\n\nMorris SA, Asnake B, Yen GG: Dendrogram seriation using simulated annealing. Information Visualization. 2003; 2(2): 95–104. Publisher Full Text\n\nJames G, Witten D, Hastie T, et al.: An Introduction to Statistical Learning, of Springer Texts in Statistics. Springer New York, New York, NY. 2013; 103. Publisher Full Text\n\nGruvaeus G, Wainer H: Two Additions to Hierarchical Cluster Analysis. J Math Stat Psychol. 1972; 25(2): 200–206. Publisher Full Text\n\nChae M, Chen JJ: Reordering hierarchical tree based on bilateral symmetric distance. PLoS One. 2011; 6(8): e22546. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHahsler M, Hornik K, Buchta C: Getting Things in Order: An Introduction to the R Package seriation. 2001. Reference Source\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\nQuackenbush J: Computational analysis of microarray data. Nat Rev Genet. 2001; 2(6): 418–27. PubMed Abstract \| Publisher Full Text\n\nBuchta C, Hornik K, Hahsler M: Getting things in order: an introduction to the R package seriation. J Stat Softw. 2008; 25(3). Reference Source\n\nR Core Team and contributors worldwide. R: Edgar Anderson’s Iris Data.\n\nWare C: Information Visualization: Perception for Design. Morgan Kaufmann Publishers Inc., San Francisco. 2004. Reference Source\n\nWare C: Color sequences for univariate maps: theory, experiments and principles. IEEE Comput Graph Appl. 1988; 8(5): 41–49. Publisher Full Text\n\nThe Cancer Genome Atlas Research Network. Comprehensive Molecular Characterization of Gastric Adenocarcinoma. Nature. 2014. Publisher Full Text\n\nTufte ER: The Visual Display of Quantitative Information. Graphics Press, Cheshire, CT, USA. 1986. Reference Source\n\nSakai R, Winand R, Verbeiren T, et al.: R package dendsort for modular leaf ordering methods. Zenodo. 2014. Data Source" }	[ { "id": "5627", "date": "22 Aug 2014", "name": "Eamonn Maguire", "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 piece of work which is immediately available to the bioinformatics community via the implementation in R and deployment to the CRAN repository.Within the visualization community, we know that heat map representations work much better when clustered appropriately since correlations with colour are more difficult to detect when colour grouping is not present, see Haroz and Whitney (2012). This technique presented by the authors goes some way to making heat map representations more effective.Their evaluation is complete and shows clear advantages over existing methods.", "responses": [] }, { "id": "5926", "date": "09 Sep 2014", "name": "Jan Oosting", "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 method to improve the interpretation of dendrograms from clustering algorithms.It is well written and the authors go into sufficient detail to describe the issues with the available ways to display dendrograms. They have applied their method to a number of case studies, and here they show that their method improves the distinction between clusters in the data.The method is made available in the form of an R package (dendsort). Creating re-ordered dendrograms is quite easy with this package, but I found it more cumbersome to apply the method to heatmaps. The standard heatmap() function has a 'reorderfun' parameter, but it expects other parameters than the dendsort() function. The documentation of the package should be improved to make this easier.", "responses": [] } ]	1	https://f1000research.com/articles/3-177
https://f1000research.com/articles/3-103/v1	09 May 14	{ "type": "Research Article", "title": "Abnormal hemostatic function one year after orthotopic liver transplantation can be fully attributed to endothelial cell activation", "authors": [ "Freeha Arshad", "Jelle Adelmeijer", "Hans Blokzijl", "Aad van den Berg", "Robert Porte", "Ton Lisman", "Freeha Arshad", "Jelle Adelmeijer", "Hans Blokzijl", "Aad van den Berg", "Robert Porte" ], "abstract": "Background: The long-term risk of thrombotic and vascular complications is elevated in liver transplant recipients compared to the general population. Patients with cirrhosis are in a hypercoagulable status during and directly after orthotopic liver transplantation, but it is unclear whether this hypercoagulability persists over time.Aim: We aimed to investigate the hemostatic status of liver transplant recipients one year after transplantation.Methods: We prospectively collected blood samples of 15 patients with a functioning graft one year after orthotopic liver transplantation and compared the hemostatic status of these patients with that of 30 healthy individuals.Results: Patients one year after liver transplantation had significantly elevated plasma levels of von Willebrand factor (VWF). Thrombin generation, as assessed by the endogenous thrombin potential, was decreased in patients, which was associated with increased plasma levels of the natural anticoagulants antithrombin and tissue factor pathway inhibitor. Plasma fibrinolytic potential was significantly decreased in patients and correlated inversely with levels of plasminogen activator inhibitor-1.Conclusion: One year after liver transplantation, liver graft recipients have a dysregulated hemostatic system characterised by elevation of plasma levels of endothelial-derived proteins. Increased levels of von Willebrand factor and decreased fibrinolytic potential may (in part) be responsible for the increased risk for vascular disease seen in liver transplant recipients.", "keywords": [ "orthotopic liver transplantation", "hemostasis", "hypercoagulability", "thrombotic complications" ], "content": "Introduction\n\nPatients with chronic liver disease frequently have major and multiple alterations in their hemostatic system, including a decreased platelet count and decreased plasma levels of pro- and anti-hemostatic proteins produced by the diseased liver1. The decrease in procoagulant proteins is evidenced by prolonged test results of routine coagulation assays such as the prothrombin time (PT) and activated partial thromboplastin time (APTT). Historically, due to an increased bleeding risk during surgery in combination with prolonged conventional coagulation tests and thrombocytopenia, liver disease patients were thought to be in a hypocoagulable state. In recent years it has become increasingly accepted that cirrhosis patients have a rebalanced hemostatic system which is not adequately represented by routine coagulation tests as they are only sensitive for procoagulant proteins and do not take the concomitant decrease in antihemostatic proteins into account2,3. The rebalanced hemostatic system is more fragile as compared to healthy individuals and may decompensate towards hypo- or hypercoagulability by factors such as renal failure, trauma, infection, and surgery1. Besides bleeding complications, patients with cirrhosis are also at risk for thrombotic complications and this particular clinical scenario has only recently been fully appreciated4–8.\n\nPatients with cirrhosis who undergo orthotopic liver transplantation (OLT) show a rapid normalisation of coagulation proteins due to the intact synthetic capacity of the transplanted liver. Previous studies performed in our laboratory have shown that hemostatic capacity early after OLT appears adequate, but shows important differences when compared to healthy individuals. At 10 days after OLT, when synthetic function of the liver as assessed by PT and APTT values is adequate, multiple laboratory parameters suggest that the patients are in a hypercoagulable state. Specifically, we have shown an unbalanced von Willebrand factor (VWF)/ADAMTS13 system9 and enhanced thrombin generation2. Also, we have shown a decreased fibrinolytic potential up to five days after surgery10.\n\nClinically, this hypercoagulable status is evidenced by a profoundly increased risk for thrombotic complications such as hepatic artery thrombosis (HAT). While previously HAT was assumed to be a solely surgical complication, there is emerging evidence for the involvement of the hemostatic system in the development of HAT11. In addition, liver transplant recipients are at increased risk for arterial thrombotic events. The risk for thrombotic complications remains increased months and even years after OLT compared to the general population and a substantial part of morbidity and mortality in liver transplant recipients who survive the first year after transplantation is due to vascular events12,13. Long-term vascular complications are mainly ascribed to the use of immunosuppressive medication12. Besides the known metabolic risk profile associated with the use of immunosuppressive medication, several in vitro studies have provided evidence for a prohemostatic effect of such drugs14,15.\n\nWhile there is laboratory evidence for a hypercoagulable state during and directly after OLT, it is unclear whether the hypercoagulability persists and, if so, for how long. To our knowledge there has been no study investigating the hemostatic potential in liver transplant recipients long after a successful transplant. We aimed to investigate the long-term status of the hemostatic system by various assays of hemostatic competence in patients one year after OLT. Understanding the hemostatic state of transplanted patients is essential for clinical practice and for the development of preventive measures for short- and long term vascular complications.\n\n\nMethods\n\nWe designed a prospective cohort study. Fifteen adult patients who visited the outpatient Hepatology Clinic of the University Medical Center Groningen (UMCG) in The Netherlands for their one-year follow-up visit after OLT, and had adequate liver function as assessed by routine laboratory parameters such as aspartate aminotransferase (ASAT) and alanine aminotransferase (ALAT), bilirubin, albumin, and PT, were included in this study. The control group consisted of 30 healthy volunteers from our laboratory staff. Patients and controls with a history of thrombotic complications, congenital coagulation disorders, active graft rejection, active infection, or who had used anticoagulant drugs in the past 10 days, suffered from disease recurrence, or were pregnant were excluded. A brief questionnaire was used to collect demographic and disease information (Supplementary File 1).\n\nBlood samples were drawn by veni-puncture and collected into vacuum tubes containing 3.8% trisodium citrate as an anticoagulant (Becton Dickinson, Breda, The Netherlands), at a blood to anticoagulant ratio of 9:1. Platelet-poor plasma was prepared by double centrifugation at 2000g and 10.000g respectively for 10 min. Plasma was snap-frozen and stored at -80°C until use.\n\nPlasma levels of VWF were determined with an in-house enzyme-linked immunosorbent assay (ELISA) using commercially available polyclonal antibodies (A0082 for coating and P0226 for detection, both are rabbit anti-human antibodies, P0226 is a horseradish-peroxidase conjugated version of A0082 (RRID:AB_579516), DAKO, Glostrup, Denmark). A disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13) activity was measured in plasma which was pretreated for 30 minutes at 37°C with bilirubin oxidase (10U/mL; Sigma-Aldrich, Zwijndrecht, The Netherlands) to avoid interference of bilirubin with the assay. ADAMTS13 activity was assessed using the FRETS-VWF73 assay (Peptanova, Sandhausen, Germany) based on method described by Kokame et al.16. The antigen levels of VWF and the activity of ADAMTS13 in pooled normal plasma were set at 100%, and values obtained in test plasmas were expressed as a percentage of pooled normal plasma.\n\nPlatelet activation was assessed by measuring plasma levels of soluble P-selectin and platelet factor 4 (PF4) with a commercially available ELISAs (R&D Systems, Abingdon, United Kingdom).\n\nThe thrombin generation test was performed using platelet-poor plasma (PPP) with the fluorimetric method described by Hemker, Calibrated Automated Thrombography® (CAT)17. Coagulation was activated using a commercial trigger composed of recombinant tissue factor (TF) at a concentration of 4 pM and phospholipids at a concentration of 4 μM, in the presence or absence of soluble thrombomodulin (TM) (Thrombinoscope BV, Maastricht, The Netherlands). Thrombin Calibrator (Thrombinoscope BV, Maastricht, The Netherlands) was added to the wells containing plasma to calibrate the thrombin generation curves. A fluorogenic substrate with CaCl2 (FluCa-kit, Thrombinoscope BV, Maastricht, The Netherlands) was dispensed in each well to allow a continuous registration of thrombin generation. Fluorescence produced was read every 20 seconds by a fluorometer, Fluoroskan Ascent® (ThermoFisher Scientific, Helsinki, Finland). All experiments were performed in triplicate.\n\nThe endogenous thrombin potential (ETP), peak height, velocity index and lag time were derived from the thrombin generation curves by the Thrombinoscope software.\n\nLevel of prothrombin F1+2 fragment in plasma were determined with a commercially available ELISA (Siemens, Breda, The Netherlands) according to the manufacturer’s instructions.\n\nLevels of factor (F) VIII, II, antithrombin (AT) and Protein C were measured on an automated coagulation analyzer (ACL 300 TOP) with reagents and protocols from the manufacturer (Recombiplastin 2G and FII depleted plasma for FII, Hemosil (R) SynthASil and FVIII depleted plasma for FVIII, Liquid Antithrombin reagent for AT, and Hemosil Protein C for Protein C measurements; Instrumentation Laboratory, Breda, the Netherlands).\n\nPlasma levels of Tissue Factor Pathway Inhibitor (TFPI) were determined with an in-house ELISA as previously described18.\n\nFibrinolytic potential was assessed using a plasma-based clot lysis assay. Lysis of a tissue factor–induced clot by exogenous tissue plasminogen activator (tPA) was studied by monitoring changes in turbidity during clot formation and subsequent lysis as described previously19. In short, 50 μL plasma was pipetted in a 96-well microtiter plate. Subsequently, a mixture containing phospholipid vesicles, tPA, tissue factor, and CaCl2, adjusted to a total volume of 50 μL by addition of HEPES (N-2-hydroxytethylpiperazine-N-2-ethanesulfonic acid) buffer (25 mM HEPES, 137 mM NaCl, 3.5 mM KCl, 3 mM CaCl2, 0.1% bovine serum albumin, pH 7.4) was added using a multichannel pipette. In a kinetic microplate reader (Versamax, Molecular Devices, Sunnyvale, CA), the optical density at 405 nm was monitored every 20 seconds at 37°C, resulting in a clot-lysis turbidity profile. Clot lysis times were derived from the clot-lysis turbidity profiles using in house-generated software. The clot lysis time was defined as the time from the midpoint of the clear to maximum turbid transition, representing clot formation, to the midpoint of the maximum turbid to clear transition, representing the lysis of the clot.\n\nPlasma levels of plasminogen activator inhibitor-1 (PAI-1) levels were determined with a commercially available ELISA (Sekisui, Stamford, USA).\n\nData are expressed as means (with standard deviations (SDs)), medians (with interquartile ranges), or numbers (with percentages) as appropriate. Means of two groups were compared by Student’s t-test or Mann-Whitney U test as appropriate. Spearman’s correlation coefficient was used to assess correlation between continuous variables. P values of 0.05 or less were considered statistically significant. GraphPad Prism (San Diego, USA) and IBM SPSS Statistics 20 (New York, USA) were used for analyses.\n\nWritten informed consent was obtained from every participant in this study. The study was approved by the local Medical Ethics Committee from the University Medical Center of Groningen (protocol number 2012.098). Study procedures were in accordance with the Helsinki Declaration of 1975.\n\n\nResults\n\nAll of the patients included in this study underwent OLT between 2011 and 2012. All patients received a full-size graft. None of the patients suffered from thrombosis prior to OLT or had postoperative thrombotic complications within the first year. Five patients suffered from diabetes mellitus at time of the blood draw, and four of these were insulin-dependent. Two of these patients had developed diabetes after OLT. There were five patients that were on platelet aggregation inhibitors (calcium carbasalate or aspirin) at the time of the blood draw. Two of these patients had coronary disease for which they had undergone coronary interventions prior to OLT. One patient had left ventricular hypertrophy and one patient had paroxysmal atrial fibrillation. The fifth patient appeared to have fragile arteries at the anastomotic site during OLT for which post-operative aspirin was started. Two patients suffered from hypertension, and two patients smoked cigarettes. Patient and background characteristics are presented in Table 1.\n\n*To convert values for hemoglobin to g/dl, multiply by 1.650. To convert values for bilirubin to mg/dl divide by 88.4.\n\nPatients had significantly higher plasma levels of the platelet-adhesive protein VWF compared to healthy controls (253% (200–323) (median (IQR)) vs. 99% (63–114), respectively, Figure 1A). The activity of ADAMTS13, the VWF-cleaving protease was comparable between patients and controls (82% (75–118) vs. 94% (85–102) respectively, Figure 1B). Plasma levels of sP-selectin were significantly elevated in patients compared to controls (28.0 pmol/L (25.0–39.0) vs. 21.0 pmol/L (18.8–25.3) respectively, Figure 1C). Levels of sP-selectin were similar in patients that were on calcium carbasalate or ascal compared to those who were not (33.0 pmol/L (20.0–41.0) vs. 28.0 pmol/L (25.0–33.0) respectively; p=0.68). However, levels of PF4 were similar among patients and controls (595 ng/ml (369–912) vs. 634 ng/ml (496–786) respectively; Figure 1D).\n\nA. Plasma levels of von Willebrand factor (VWF) in patients and healthy controls. B. ADAMTS13 activity in plasma from patients and healthy controls. C. Plasma levels of soluble P-selectin in patients and controls. D. Plasma levels of Platelet Factor 4 in patients and controls. Horizontal bars indicate medians.\n\nThrombin generation assays showed that patients had a decreased procoagulant capacity, both in presence and absence of thrombomodulin (Figure 2). Specifically, patients had a decreased ETP compared to controls, both in presence and absence of thrombomodulin (344 nM IIa×min (284–414) vs. 492 nM IIa×min (385–693) respectively in presence of thrombomodulin). Patients also had a decreased peak height and velocity index, and a prolonged lagtime compared to controls (Table 2).\n\nHorizontal bars indicate medians.\n\nData are presented as medians with interquartile range.\n\nThe ETP ratio, an index of the anticoagulant capacity of the protein C system defined as the ratio of the ETP with-to-without TM, was significantly lower in patients compared to controls (Table 2).\n\nPlasma levels of FII were similar in patients and controls (99 % (94–111) vs. 106% (96–117) respectively, Figure 3A). Levels of FVIII on the other hand were significantly higher in patients compared to controls (122% (111–153) vs. 87% (74–109) respectively, Figure 3B).\n\nA. Plasma levels of factor II in patients and healthy controls. B. Factor VIII plasma levels in patients and healthy controls. Horizontal bars indicate medians.\n\nPlasma levels of TFPI were significantly higher in patients compared to controls (184% (147–204) vs. 127% (82–148) respectively, Figure 4A). There was no difference in protein C levels between the groups (107% (87–124) vs. 104% (95–126) respectively, Figure 4B). In the patient group protein C correlated inversely with ETP in presence of TM (Figure 4C). Levels of AT were slightly, but significantly higher in patients compared to controls (114% (99–134) vs. 104% (97–113) respectively, Figure 4D).\n\nA. Plasma levels of TFPI in patients and healthy controls. B. Protein C levels in plasma from patients and healthy controls. C.Correlation between the Endogenous Thrombin Potential (ETP) measured in the presence of thrombomodulin (TM) and plasma levels of protein C in patients. D. Antithrombin (AT) levels in plasma from patients and healthy controls. Horizontal bars indicate medians.\n\nPlasma levels of prothrombin fragment 1+2, an indicator of in vivo thrombin generation, were similar between patients and controls (216 pmol/L (146–260) vs. 178 pmol/L (136–210) respectively, Figure 5).\n\nHorizontal bars indicate medians.\n\nClot lysis times were significantly prolonged in patients compared to controls (66.8 min (61.3–75.1) vs. 54.2 min (50.1–60.8) respectively Figure 6A and B). Plasma levels of PAI-1 were significantly higher in patients compared to controls (8.2 ng/ml (4.5–11.8) vs. 2.1 ng/ml (2.6–5.4) respectively) and correlated with clot lysis time (Figure 6C and D).\n\nA.Clot lysis time assessed in plasma from patients and healthy controls. B. Plasma levels of PAI-1in patients and healthy controls. C. Correlation between clot lysis times and PAI-1 plasma levels in patients and D. controls. Horizontal bars in panels A and B indicate medians.\n\n\nDiscussion\n\nThe main finding of this study is that one year after OLT patients with a well- functioning graft are in a hypercoagulable state. This hypercoagulable state is caused by elevated plasma levels of VWF (resulting in a hyperactive primary hemostatic system), and a hypofibrinolytic state due to elevated plasma levels of PAI-1. Nevertheless, thrombin generation was decreased in patients on year after OLT, which was related to increased levels of TFPI and AT. This decreased thrombin generation may, in part, compensate for the hypercoagulable changes. High VWF and decreased plasma fibrinolytic potential, however, are established risk factors for venous and arterial thrombosis, and we therefore speculate that the unbalanced hemostasis in patients one year after OLT may contribute to their increased risk for thrombotic events. The observed increased levels of VWF, sP-selectin, FVIII, TFPI, and PAI-1, in patients one year after OLT, are likely the result of chronic endothelial injury. All of these coagulation proteins are produced by endothelial cells. It has been demonstrated that the use of immunosuppressive drugs leads to endothelial cell activation and release of VWF14,20–25. Also, an inhibitory effect of immunosuppressive medication on fibrinolysis has been demonstrated, as evidenced by elevated levels of PAI-125–27.\n\nIt thus appears plausible that the elevation in levels of VWF, sP-selectin, FVIII, TFPI, and PAI-1, which explain the majority of the hemostatic unbalance at one year after OLT, is related to immunosuppression. Nevertheless, other causes for endothelial activation such as (de novo) diabetes, smoking, and infection or a proinflammatory status may also contribute to endothelial cell activation. Of note, none of the other drugs used by some of the patients, such as proton pump inhibitors, blood-glucose lowering medication or calcium antagonists are known to cause endothelial cell activation.\n\nPreviously we have summarized clinical and laboratory evidence for hypercoagulability as a contributor to thrombotic complications after liver transplantation11. Several of the hemostatic abnormalities that we have described in the present study have been linked to clinical thrombotic events in non-transplant patients and, therefore, are in line with our previously formulated hypotheses. Increased levels of VWF and FVIII have been (independently and in combination) described as a risk factor for venous thromboembolism but also for thromboembolic cardio- and cerebrovascular disease and mortality in several studies28–31. Hypofibrinolysis, as defined by prolonged plasma-based clot lysis times or by increased plasma levels of PAI-1 have been associated with a risk for venous thrombosis and thromboembolic cardiovascular disease32–36.\n\nWith increasing long term survival after OLT, thromboembolic cardio- and cerebrovascular disease has become an increasingly threatening factor for OLT recipients, warranting preventative measures11. While there are no laboratory studies investigating the hemostatic status of patients 5–10 years after their OLT, it is plausible that the hemostatic abnormalities induced by immunosuppressive drugs described in this study persist or even aggravate over time. The findings of this study may point to the necessity of antihemostatic treatment to prevent cardiovascular disease after OLT. While both sP-selectin and PF4 are considered markers of platelet activation, only PF4 is solely released by activated platelets. P-selectin is present in both endothelial cells as well as platelets, and a soluble fragment is released upon activation of these cells. The fact that only levels of sP-selectin were elevated in liver transplant recipients while levels of PF4 were similar to controls doesn’t indicate an increased platelet activity but rather increased endothelial cell activation. However, the elevated levels of VWF may lead thromboembolic events, which may be prevented by anti-platelet therapy. It has been demonstrated in a single retrospective study that long-term administration of aspirin lowers the incidence of HAT after OLT without increasing bleeding events37. Platelet inhibition after OLT by aspirin might not only decrease the risk for HAT but also the risk for thromboembolic cardiovascular disease similar to that in the general population, although this has not been assessed in clinical studies.\n\nThe results of our study suggest that the hemostatic imbalance of liver transplant recipients is not due to transplant-related effects, but to the endothelial activating properties of immunosuppression. Our results therefore may be likely extended to other forms of solid organ transplantation. Indeed, kidney transplant recipients are also at risk for thrombotic events, which may also in part be related to dysregulated hemostasis38–41. Although antihemostatic therapy may be beneficial, partial or complete withdrawal of immune suppression may be an optimal strategy to avoid post transplant thrombotic complication.\n\nIn conclusion, one year after liver transplantation liver transplant recipients display dysregulated hemostasis which appears to be related to endothelial activation. Whereas elevated levels of VWF and decreased fibrinolytic capacity may be related to thrombotic complications in liver transplant recipients, this risk may be attenuated in part by decreased thrombin generating capacity.\n\n\nData availability\n\nfigshare: Database containing demographic data of each patient and laboratory data of each patient and control. Doi: 10.6084/m9.figshare.100206542\n\n\nInformed consent\n\nWritten informed consent was obtained from participant in this study.", "appendix": "Author contributions\n\n\n\nTL: designed research, interpreted data, and drafted manuscript. FA: responsible for patient inclusion, interpreted data and drafted manuscript. JA: carried out laboratory measurements. HB and AvdB: responsible for patient inclusion, interpreted data. RJP: designed research, interpreted data. 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\nReferences\n\nLisman T, Porte RJ: Rebalanced hemostasis in patients with liver disease: evidence and clinical consequences. Blood. 2010; 116(6): 878–85. PubMed Abstract \| Publisher Full Text\n\nLisman T, Bakhtiari K, Pereboom IT, et al.: Normal to increased thrombin generation in patients undergoing liver transplantation despite prolonged conventional coagulation tests. J Hepatol. 2010; 52(3): 355–61. PubMed Abstract \| Publisher Full Text\n\nTripodi A, Salerno F, Chantarangkul V, et al.: Evidence of normal thrombin generation in cirrhosis despite abnormal conventional coagulation tests. Hepatology. 2005; 41(3): 553–8. PubMed Abstract \| Publisher Full Text\n\nSenzolo M, Sartori MT, Lisman T: Should we give thromboprophylaxis to patients with liver cirrhosis and coagulopathy? HPB (Oxford). 2009; 11(6): 459–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSogaard KK, Horvath-Puho E, Gronbaek H, et al.: Risk of venous thromboembolism in patients with liver disease: a nationwide population-based case-control study. Am J Gastroenterol. 2009; 104(1): 96–101. PubMed Abstract \| Publisher Full Text\n\nGulley D, Teal E, Suvannasankha A, et al.: Deep vein thrombosis and pulmonary embolism in cirrhosis patients. Dig Dis Sci. 2008; 53(11): 3012–7. PubMed Abstract \| Publisher Full Text\n\nNorthup PG, McMahon MM, Ruhl AP, et al.: Coagulopathy does not fully protect hospitalized cirrhosis patients from peripheral venous thromboembolism. Am J Gastroenterol. 2006; 101(7): 1524–8. PubMed Abstract \| Publisher Full Text\n\nGarca-Fuster MJ, Abdilla N, Fabi MJ, et al.: [Venous thromboembolism and liver cirrhosis]. Rev Esp Enferm Dig. 2008; 100(5): 259–262. PubMed Abstract\n\nPereboom IT, Adelmeijer J, van Leeuwen Y, et al.: Development of a severe von Willebrand factor/ADAMTS13 dysbalance during orthotopic liver transplantation. Am J Transplant. 2009; 9(5): 1189–96. PubMed Abstract \| Publisher Full Text\n\nLisman T, Leebeek FW, Meijer K, et al.: Recombinant factor VIIa improves clot formation but not fibrolytic potential in patients with cirrhosis and during liver transplantation. Hepatology. 2002; 35(3): 616–21. PubMed Abstract \| Publisher Full Text\n\nArshad F, Lisman T, Porte RJ: Hypercoagulability as a contributor to thrombotic complications in the liver transplant recipient. Liver Int. 2013; 33(6): 820–7. PubMed Abstract \| Publisher Full Text\n\nDuffy JP, Hong JC, Farmer DG, et al.: Vascular complications of orthotopic liver transplantation: experience in more than 4,200 patients. J Am Coll Surg. 2009; 208(5): 896–903, discussion 903–5. PubMed Abstract \| Publisher Full Text\n\nBorg MA, van der Wouden EJ, Sluiter WJ, et al.: Vascular events after liver transplantation: a long-term follow-up study. Transpl Int. 2008; 21(1): 74–80. PubMed Abstract \| Publisher Full Text\n\nHuang LQ, Whitworth JA, Chesterman CN: Effects of cyclosporin A and dexamethasone on haemostatic and vasoactive functions of vascular endothelial cells. Blood Coagul Fibrinolysis. 1995; 6(5): 438–45. PubMed Abstract \| Publisher Full Text\n\nBombeli T, Müller M, Straub PW, et al.: Cyclosporine-induced detachment of vascular endothelial cells initiates the intrinsic coagulation system in plasma and whole blood. J Lab Clin Med. 1996; 127(6): 621–34. PubMed Abstract \| Publisher Full Text\n\nKokame K, Nobe Y, Kokubo T, et al.: FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol. 2005; 129(1): 93–100. PubMed Abstract \| Publisher Full Text\n\nRuitenbeek K, Meijers JC, Adelmeijer J, et al.: Intact thrombomodulin-mediated regulation of fibrinolysis during and after liver transplantation, despite a profoundly defective thrombomodulin-mediated regulation of coagulation. J Thromb Haemost. 2010; 8(7): 1646–9. PubMed Abstract \| Publisher Full Text\n\nMaurissen LF, Castoldi E, Simioni P, et al.: Thrombin generation-based assays to measure the activity of the TFPI-protein S pathway in plasma from normal and protein S-deficient individuals. J Thromb Haemost. 2010; 8(4): 750–8. PubMed Abstract \| Publisher Full Text\n\nLisman T, de Groot PG, Meijers JC, et al.: Reduced plasma fibrinolytic potential is a risk factor for venous thrombosis. Blood. 2005; 105(3): 1102–5. PubMed Abstract \| Publisher Full Text\n\nJeanmart H, Malo O, Carrier M, et al.: Comparative study of cyclosporine and tacrolimus vs newer immunosuppressants mycophenolate mofetil and rapamycin on coronary endothelial function. J Heart Lung Transplant. 2002; 21(9): 990–8. PubMed Abstract \| Publisher Full Text\n\nWeis M, Wildhirt SM, Schulze C, et al.: Impact of immunosuppression on coronary endothelial function after cardiac transplantation. Transplant Proc. 1998; 30(3): 871–2. PubMed Abstract\n\nOvuworie CA, Fox ER, Chow CM, et al.: Vascular endothelial function in cyclosporine and tacrolimus treated renal transplant recipients. Transplantation. 2001; 72(8): 1385–8. PubMed Abstract \| Publisher Full Text\n\nSchrama YC, van Dam T, Fijnheer R, et al.: Cyclosporine is associated with endothelial dysfunction but not with platelet activation in renal transplantation. Neth J Med. 2001; 59(1): 6–15. PubMed Abstract \| Publisher Full Text\n\nMorris ST, McMurray JJ, Rodger RS, et al.: Endothelial dysfunction in renal transplant recipients maintained on cyclosporine. Kidney Int. 2000; 57(3): 1100–6. PubMed Abstract \| Publisher Full Text\n\nBaas MC, Gerdes VE, Ten Berge IJ, et al.: Treatment with everolimus is associated with a procoagulant state. Thromb Res. 2013; 132(2): 307–11. PubMed Abstract \| Publisher Full Text\n\nPatrassi GM, Sartori MT, Rigotti P, et al.: Reduced fibrinolytic potential one year after kidney transplantation. Relationship to long-term steroid treatment. Transplantation. 1995; 59(10): 1416–20. PubMed Abstract\n\nLevi M, Wilmink J, Büller HR, et al.: Impaired fibrinolysis in cyclosporine-treated renal transplant patients. Analysis of the defect and beneficial effect of fish-oil. Transplantation. 1992; 54(6): 978–83. PubMed Abstract\n\nTsai AW, Cushman M, Rosamond WD, et al.: Coagulation factors, inflammation markers, and venous thromboembolism: the longitudinal investigation of thromboembolism etiology (LITE). Am J Med. 2002; 113(8): 636–42. PubMed Abstract \| Publisher Full Text\n\nMartinelli I: von Willebrand factor and factor VIII as risk factors for arterial and venous thrombosis. Semin Hematol. 2005; 42(1): 49–55. PubMed Abstract \| Publisher Full Text\n\nFolsom AR, Rosamond WD, Shahar E, et al.: Prospective study of markers of hemostatic function with risk of ischemic stroke. The Atherosclerosis Risk in Communities (ARIC) Study Investigators. Circulation. 1999; 100(7): 736–42. PubMed Abstract \| Publisher Full Text\n\nJansson JH, Nilsson TK, Johnson O: von Willebrand factor, tissue plasminogen activator, and dehydroepiandrosterone sulphate predict cardiovascular death in a 10 year follow up of survivors of acute myocardial infarction. Heart. 1998; 80(4): 334–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlmér LO, Ohlin H: Elevated levels of the rapid inhibitor of plasminogen activator (t-PAI) in acute myocardial infarction. Thromb Res. 1987; 47(3): 335–9. PubMed Abstract \| Publisher Full Text\n\nMeltzer ME, Lisman T, de Groot PG, et al.: Venous thrombosis risk associated with plasma hypofibrinolysis is explained by elevated plasma levels of TAFI and PAI-1. Blood. 2010; 116(1): 113–21. PubMed Abstract \| Publisher Full Text\n\nMeltzer ME, Lisman T, Doggen CJ, et al.: Synergistic effects of hypofibrinolysis and genetic and acquired risk factors on the risk of a first venous thrombosis. PLoS Med. 2008; 5(5): e97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamsten A, Wiman B, de Faire U, et al.: Increased plasma levels of a rapid inhibitor of tissue plasminogen activator in young survivors of myocardial infarction. N Engl J Med. 1985; 313(25): 1557–63. PubMed Abstract \| Publisher Full Text\n\nReddel CJ, Curnow JL, Voitl J, et al.: Detection of hypofibrinolysis in stable coronary artery disease using the overall haemostatic potential assay. Thromb Res. 2013; 131(5): 457–62. PubMed Abstract \| Publisher Full Text\n\nVivarelli M, La Barba G, Cucchetti A, et al.: Can antiplatelet prophylaxis reduce the incidence of hepatic artery thrombosis after liver transplantation? Liver Transpl. 2007; 13(5): 651–4. PubMed Abstract \| Publisher Full Text\n\nIrish AB, Green FR: Environmental and genetic determinants of the hypercoagulable state and cardiovascular disease in renal transplant recipients. Nephro Dial Transplant. 1997; 12(1): 167–73. PubMed Abstract \| Publisher Full Text\n\nAllen RD, Michie CA, Murie JA, et al.: Deep venous thrombosis after renal transplantation. Surg Gynecol Obstet. 1987; 164(2): 137–42. PubMed Abstract\n\nHumar A, Johnson EM, Gillingham KJ, et al.: Venous thromboembolic complications after kidney and kidney-pancreas transplantation: a multivariate analysis. Transplantation. 1998; 65(2): 229–34. PubMed Abstract\n\nThe 12th Report of the Human Renal Transplant Registry. Prepared by the Advisory Committee to the Renal Transplant Registry. JAMA. 1975; 233(7): 787–96. PubMed Abstract \| Publisher Full Text\n\nArshad F, Adelmeijer J, Blokzijl H, et al.: Database containing demographic data of each patient and laboratory data of each patient and control. Figshare. 2014. Data Source" }	[ { "id": "4734", "date": "05 Jun 2014", "name": "Hugo Ten Cate", "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\nLisman and colleagues follow up on a number of prior studies of hemostatic and coagulation activity status in patients with liver cirrhosis. In this small study they document the hemostatic phenotype in plasma from 15 patients that had underwent orthotopic liver transplantation (OLT) for liver cirrhosis of diverse nature. In the cross sectional study they compare with healthy persons (of comparable age and sex?).The main finding is that as compared to healthy persons the 15 patients had elevated plasma levels of von Willebrand factor (vWF), plasminogen activator inhibitor-1 (PAI-1), FVIII and soluble P selection, suggestive of an activated endothelium (and/or activated hemostasis) and an attenuated fibrinolytic system, as shown with clot lysis methodology. Interestingly, thrombin generation parameters were either not different (F1+2) or lowered in case of the potential to form thrombin (measured by Calibrated Automated Thrombinography). The authors conclude that there is evidence of a hypercoagulable state at one year after OLT in these patients, given the above changes, which are suggestive of endothelial activation or damage. A major contributing factor in this process may be the use of immunosuppresive agents. In addition they document a hypofibrinolytic state.Minor comments:Obviously, this study cannot give insight in the origin of the found abnormalities and it is impossible to dissect effects of immunosuppression from those of underlying morbidity; there are several risk factors for CV disease present in this population that may play a role in endothelial activation. A question would be whether any of these patients had been characterised prior to OLT, since it is known that many hemostatic abnormalities are present in association with liver cirrhosis, as the authors also have documented in previous publications. For instance the high level of vWF must (likely) have been present before surgery so that it remains a question how much of the present findings are due to immunosuppression, or to prior comorbidity? Also, the comparison with healthy persons from the lab (age? sex?) may suggest that differences in markers are substantial while in part these may be due to young age or healthy status in controls, so the differences attributable to OLT and/or immunosuppression may be smaller when compared to a comparable group of controls, rather than healthy laboratory subjects. Interestingly, indeed the hemostatic activation may be counterbalanced by a lowered thrombin generation activity, at least by CAT, so that the net hyper coagulability is somewhat blunted. Given somewhat elevated levels of AT and TFPI, these may explain the lowered ETP and peak levels, but there is no direct evidence that these are the major modifiers in this case. In summary, this is a small but interesting study, suggesting a prothrombotic plasma phenotype based on activated endothelial markers coupled to a hypofibrinolytic state, related to elevated PAI-1. Prospective studies of the actual changes pre and post OLT may be needed to get a better idea of the contribution of comorbidity to these changes, preferably as compared to individuals of similar sex and age. The relevance of the observed changes for predicting the risk of thrombosis in these vulnerable patients is worth pursuing.", "responses": [ { "c_id": "913", "date": "22 Jul 2014", "name": "Ton Lisman", "role": "Author Response F1000Research Advisory Board Member", "response": "We appreciate the interest of the reviewer in our study and thank him for his constructive comments. We agree that our study is not able to differentiate between the effects of immunosuppression and the effects of comorbidities and smoking on levels of endothelial-derived proteins. We have adjusted the conclusion section to indicate that both immunosuppression and comorbidities may be responsible for the increased levels of endothelial-derived proteins in transplant recipients. The patients were not characterized prior to liver transplantation. Indeed, patients with liver disease have multiple alterations in their hemostatic system which are attributed to defective hepatic synthesis, defective hepatic clearance, consumption, and endothelial activation (which in patients with cirrhosis is strongly related to portal hypertension). Many of the disease-related changes in the hemostatic system are likely no longer present at one year after transplantation in patients with a well-functioning graft. Defects in synthesis and clearance are no longer present in patients with a functioning graft, and also portal hypertension will be no longer present. We thus expect the hemostatic changes observed in the present study not to be related to the pre-transplant liver disease, in particular since patients that had disease recurrence were not included in our study. As indicated in our response to the comments of Dr. Gatt, our controls were purposely not matched for age and sex (now indicated in the materials and methods section), and were indeed much younger than the age of our patients. We agree that some of the markers are age-dependent, but the increases observed in endothelial-derived proteins are much larger than expected based on age differences only. We nevertheless agree with the reviewer that the differences may be slightly overestimated as a result of the age difference. Although we agree that we did not directly show that the decreased thrombin generation in our patients was related to increased TFPI and AT levels, there is evidence that these 2 proteins are important modifiers of the thrombin generation test. We have added a reference to a paper showing the importance of TFPI and AT in determining thrombin generation." } ] }, { "id": "5186", "date": "23 Jun 2014", "name": "Annabel Blasi", "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 study, the authors investigated the hemostatic status of liver transplant recipients one year after transplantation, finding a prothrombotic profile based on activated endothelial markers coupled to a hypofibrinolytic state. It point out a very relevant clinical problem as it is the thrombotic complications in the post transplant period. The aim and methods are well defined and well performed, however I would like the authors address some aspects about the study population:The study population shows several factors which have clearly impact on the endothelial function, apart from immunosuppressive therapy. In order to obtain more information from the data, I suggest that the authors discuss the differences, if any, about the more relevant results: VW, VIII, sP-selectin, TFPI and PAI-1 according to the following sub-groups of patients:DiabetesCardiovacular diseaseEtiology of liver disease - Given that a hypercoagulable/hypofibrinolytic state has been suggested in several of the reported etiologies in the study group (biliary cirrhoris, NASH, familial amyloidotic polyneuropathy) those reversibilities after liver transplantation have not been proven.Immunosuppressive regiment.To discard that some of the aforementioned conditions confers a risk extra for thrombosis. Otherwise, patients had a decreased ETP compared to controls, both in presence and absence of thrombomodulin. The ETP ratio, an index of the anticoagulant capacity of the protein C system defined as the ratio of the ETP with-to-without thrombomodulin was significantly lower in patients compared to controls but interestingly, there was no difference in protein C levels between both groups. Five patients were on platelet aggregation inhibitors. Levels of sP-selectin were similar in patients that were on aspirin compared to those who were not. Could the authors make some comment about this? Aspirin resistance in this population could be suggested? Although the size and the characteristics of the study population entail some limitations, it is a good starting point (never addressed before), for further studies with a larger sample.", "responses": [ { "c_id": "911", "date": "22 Jul 2014", "name": "Ton Lisman", "role": "Author Response F1000Research Advisory Board Member", "response": "We appreciate the interest of the reviewer in our study and thank her for her constructive comments.The suggestion to look at subgroups is appreciated, and we agree that comorbidities such as diabetes and cardiovascular disease may in itself have an effect on levels of endothelial-derived proteins. In addition, we agree that the etiology of liver disease prior to transplant may be of interest (although the patients were only included when they did not have evidence of disease recurrence), and that the different immunosuppressive regimens may have different effects on levels of endothelial-derived proteins. However, we feel that the small size of our study doesn’t allow for a meaningful subgroup analysis. The normal ranges of the proteins and tests studied are wide already in the general population (also evidenced by the ranges observed in our 30 healthy controls), which indicates that subgroups should be reasonably sized to allow a meaningful subgroup analysis. We agree that the similar P-selectin levels in the patients that were on aspirin therapy may indicate aspirin resistance, although we did not assess platelet function directly and the group of patients on aspirin was small. It has to be noted, however, that the circulating P-selectin is in part also released from endothelial cells, and this process is not expected to be influenced by aspirin. Nevertheless, liver transplant recipients often have multiple factors (diabetes, smoking, hyperlipidemia that may increase the risk for aspirin resistance. To our knowledge there are no published data on the frequency of aspirin resistance in liver transplant recipients or patients with liver disease and such studies will be required to optimize antihemostatic therapy in these patients." } ] }, { "id": "5188", "date": "04 Jul 2014", "name": "Alex Gatt", "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\nArshad et al. present an interesting study on the haemostatic changes seen 1 year after liver transplantation. This is of particular interest to hepatologists, liver surgeons and haemostasiologists alike. The Groningen group clearly show that these patients, even 1 year after transplantation, still have high FVIII, and VWF. However they do recover their anticoagulant proteins, leading to a brisk reduction of thrombin generation upon the addition of thrombomodulin.This is a proof of concept study due to the small number of patients. There is a major confounder, as the authors themselves acknowledge in their introduction, which is the chronic immunosuppression which could be leading this 'endothelial activation', but nevertheless whatever the cause it looks likely that this is present and might play a part in the post transplant thrombotic events in these patients. Ideally, the authors should specify what Protein C and AT assays they used, i.e. are they activity assays? Another confounder is that some patients were on antiplatelet agents and other were not. It is not clear why the patients still had lower thrombin generation without thrombomodulin. This could be related to the concentration of TF used in the assay. Were the controls age and sex matched? Finally, despite the fact that VWF, FVIII and PAI-1 are still elevated 1 year after transplantation, looking at the TG data, the patients seem to be hypocoagulable with lower Velocity Index, prolonged LT and lower ETP. There is also a seemingly more profound effect of the natural anticoagulants upon addition of TM. So is this still a case of balanced haemostasis?", "responses": [ { "c_id": "912", "date": "22 Jul 2014", "name": "Ton Lisman", "role": "Author Response F1000Research Advisory Board Member", "response": "We appreciate the interest of the reviewer in our study and thank him for his constructive comments.The Protein C and AT assays are indeed activity assays using chromogenic substrates. As indicated in the manuscript, these assays were purchased from Instrumentation Laboratory. For a detailed description of the assays see: http://benelux.instrumentationlaboratory.com/products-services/hemostasis-diagnostics/reagents/thrombophilia.aspx. Although we agree that the use of antiplatelet agents may be a confounder (although antiplatelet agents to our knowledge have not been shown to alter the outcome of most of the assays we have used), we would like to stress that the clinical reality is that liver transplant recipients are frequently prescribed multiple drugs other than immunosuppressants. We were interested to find out whether coagulation abnormalities may contribute to the increased thrombotic risk after liver transplantation. Thus, while acknowledging the fact that the causes of the hemostatic changes observed in liver transplant recipients are likely multicausal (and not fully elucidated by our study), our data show evidence for a hypercoagulable profile in patients one year after transplantation. The reduced thrombin generation in the absence of thrombomodulin is likely due to the elevated levels of TFPI and AT, which have been shown to be key determinants of thrombin generation (J Thromb Haemost. 2008 Jan;6(1):125-31). This reference has been added to the manuscript. Controls were not age and sex matched. We included controls to establish reference values for the various assays performed in the study, similar to how reference values in routine diagnostic laboratories are established. While acknowledging that not all measured variables are constant with age, the differences observed in the endothelial-derived proteins are substantial and not explained by minor changes caused by age. We have altered the text in the materials and methods section to indicate that controls were included only to establish reference ranges, and included age and sex of the controls. We agree that we find evidence of both hyper- and hypocoagulable changes in the liver transplant recipients. Our conclusion thus is not that the hemostatic system is in balance, as suggested by the reviewer. Rather, we conclude there is dysregulated hemostasis. The final sentence of the conclusion sums up how we interpret our findings of the study: “Whereas elevated levels of VWF and decreased fibrinolytic capacity may be related to thrombotic complications in liver transplant recipients, this risk may be attenuated in part by decreased thrombin generating capacity.”" } ] } ]	1	https://f1000research.com/articles/3-103
https://f1000research.com/articles/2-96/v1	28 Mar 13	{ "type": "Review", "title": "Intrathecal Drug Delivery (ITDD) systems for cancer pain", "authors": [ "Gaurav Bhatia", "Mary E Lau", "Padma Gulur", "Gaurav Bhatia", "Mary E Lau" ], "abstract": "Intrathecal drug delivery is an effective pain management option for patients with chronic and cancer pain. The delivery of drugs into the intrathecal space provides superior analgesia with smaller doses of analgesics to minimize side effects while significantly improving quality of life. This article aims to provide a general overview of the use of intrathecal drug delivery to manage pain, dosing recommendations, potential risks and complications, and growing trends in the field.", "keywords": [ "Intrathecal drug delivery", "chronic pain", "cancer", "analgesic", "opioids" ], "content": "Introduction\n\nIntrathecal drug delivery (ITDD or IDD) is a method which provides analgesia by a direct continuous infusion of medication into the cerebrospinal fluid (CSF). The rationale is based on the concept that direct delivery into the CSF within the intrathecal space avoids crossing the blood-brain barrier and allows delivery of smaller doses than oral, intravenous, transdermal or epidural drug administration1. The spinal cord is a key anatomical site for pain processing and more specifically the dorsal horn of the spinal cord within the intrathecal space. There are various receptor targets within the spinal column that can be utilized in relieving pain. These receptors include opioid (mu, kappa, and delta), GABA, alpha-2, dopaminergic, NMDA as well as sodium and calcium channels2. The advantage of delivering the medication directly into the CSF in smaller dosages to target the specific receptors is that it may reduce the systemic side effects and the direct delivery to the site of action allows a more rapid and effective response. As a popular alternative for patients experiencing cancer or chronic pain, intrathecal analgesia has become a therapeutic option for patients who have exhausted all other treatment avenues as well as patients experiencing side effects from their current treatment options. The system uses a small pump that is surgically placed under the skin of the abdomen and delivers medication through a tunneled catheter directly into the intrathecal space3. The pump is refilled periodically through a subcutaneous access port in an office setting utilizing a sterile technique. The pump can be programmed to be a fixed-rate, constant flow device or a variable rate pump with the option of boluses via patient controlled analgesia programming using a wireless radiofrequency transmitter given to the patient1. In contrast to more conservative oral or intravenous delivery methods, ITDD is becoming increasingly popular for its efficacy in managing cancer and chronic pain. This review will specifically target ITDD uses for patients with cancer-related pain although its use has also been very popular in patients with chronic non-cancer pain.\n\n\nCancer pain: epidemiology\n\nThe incidence of deaths due to cancer has been estimated to be 6.6 million people globally each year4. Pain is often considered as the most feared symptom amongst individuals living with cancer and it can occur at any point during the course of the illness. Pain is often also the first sign of malignancy4. Pain is also often associated with the treatment of cancer and it can present itself as both acute and chronic pain. Furthermore, advancement of the disease process can lead to pain or make the existing pain worse. It is difficult to estimate the prevalence of cancer pain because of a lack of standardization in definitions of pain and in the measures used to assess it5. The other factor that makes it challenging in estimating the prevalence is the heterogeneity of nociceptive and neuropathic pain conditions6. About 50% of patients will report pain as a symptom at the time of cancer diagnosis and early in the course of disease6. Its prevalence increases to 75% or more at advanced stages. A strategy should be devised to evaluate the prevalence of pain in patients with cancer by separating the pain related to the disease process, to its treatment, or to unrelated causes.\n\n\nPrevalence of pain related to cancer\n\nAccording to a systematic review of the past 40 years by van den Beuken et al. published in the Annals of Oncology in 20074, the range of reported prevalence of pain is highest for the following tumors:\n\nHead and neck (67–91%)\n\nProstate (56–94%)\n\nUterine (30–90%)\n\nGenitourinary (58–90%)\n\nBreast (40–89%)\n\nPancreatic (72–85%)\n\n\nTreatment options\n\nThe various treatment options for cancer pain generally depend on the location, character, duration, its origin, and the success or failure of previous and ongoing treatments. Treatment options for cancer-related pain include oral non-steroidal anti-inflammatory drugs, oral opioid analgesics, neuropathic pain medications, parenteral opioids, peripheral nerve blocks, neurolysis such as celiac plexus blocks, local anesthetic injections, spinal cord stimulation, continuous epidural analgesia, and intrathecal pumps. In most cases, multiple forms of treatment are used for the most effective treatment combination. Opiates such as morphine, dilaudid, fentanyl, and sufentanil are most commonly administered via the intrathecal drug delivery system for cancer-related and non-malignant pain7. There is increasing evidence that intrathecal opioids are superior to oral delivery in malignant pain, especially when narcotic dosage is limited by its side effects1. It is believed that this advantage results from delivering extremely small doses on target receptors in the spinal column8 (see Table 1).\n\nAll other drugs are converted to Morphine Equivalent Dose (MED) to allow equianalgesic conversions.\n\nThere are other agents also available to treat neuropathic pain, spasticity related pain, sympathetic pain, and visceral pain (see Table 2 for examples). In addition, ITDD systems can also be used to deliver chemotherapy agents such as floxuridine and methorexate for the treatment of primary or metastatic cancer. Table 2 below lists the most commonly targeted receptors and the medications used for each receptor.\n\nNote: The only medications that are FDA approved for ITDD are: Morphine Sulfate, Baclofen, Ziconotide, Floxuridine, and Methotrexate9.\n\nOne of the most important clinical decisions a provider has to make is what agent to use in the intrathecal pump and in some cases, multiple forms of treatment are used for the most effective treatment combination. In this review, we looked at the recommendations of The Polyanalgesic Consensus Conference (PACC) 2012 which lists the algorithms for intrathecal therapies in both neuropathic and nociceptive pain (Table 3 and Table 4). The algorithm lists the medications that are arranged in a hierarchy on the basis of evidence of efficacy and safety. The approaches are listed from Line 1 (first line approach) to Line 5 (a more advanced approach if previous approaches are unsuccessful).\n\nNote: Line 6- Experimental agents: Gabapentin, Octreotide, Conpeptide, Neostigmine, Adenosine, XEN 2174, AM 336, XEN, ZGX 160 (ongoing experiments in animal models).\n\nNote: Line 6- Experimental agents: Gabapentin, Octreotide, Conpeptide, Neostigmine, Adenosine, XEN 2174, AM 336, XEN, ZGX 160 (ongoing experiments in animal models).\n\nA retrospective study10 involving 120 patients with non-malignant pain syndromes found that the patients benefited from the intrathecal opiate therapy and the average pain reduction was 67.4% after 6 months from implantation. Over ninety percent of the patients were satisfied with the therapy and 81% reported an improvement in their quality of life.\n\nBesides opiates, other medications such as baclofen (GABA agonist) for spasticity related pain, ziconotide (calcium channel blocker) for severe chronic neuropathic pain, bupivicaine (amide local anesthetic), clonidine (alpha-2 agonist), and ketamine (NMDA antagonist) have also been used in various instances and in specific patient groups. In addition, ITDD systems can also be used to deliver chemotherapy agents such as floxuridine and methorexate for the treatment of primary or metastatic cancer9.\n\nOne of the most common uses of the ITDD system is the delivery of baclofen for spasticity related pain3. Baclofen is a GABA-A receptor agonist which can have prominent effect on the motor tone via direct hyperpolarization of the motor horn cells. It is most often utilized in patients suffering from spasticity, cerebral palsy, amyotrophic lateral sclerosis, Stiff-Pearson syndrome, and those suffering from brain and spinal cord injuries7.\n\n\nIntrathecal drug delivery systems\n\nOne option, intrathecal analgesia, has become a therapeutic option for patients who have exhausted all other treatment avenues as well as patients experiencing side effects from their current treatment options3. In intrathecal drug delivery or ITDD, medication is delivered directly into the CSF in the intrathecal space (Figure 1); therefore, a substantially smaller dose of medication is required compared to the oral or intravenous route. With lesser dosage of the medication and a more direct route to the pain processing center in the spinal cord, the patient generally experiences superior analgesia with fewer side effects such as nausea, pruritis, erythema, flushing, constipation, etc. However, intrathecal delivery of medication also poses significant risks such as respiratory depression, infection, bleeding, epidural/spinal hematomas, spinal cord injury during initial catheter placement, wound dehiscence, pocket seromas, pump malfunction, and costs.\n\nReprinted with the kind permission of Medtronic, Inc.\n\nThere have been various publications and consensus guidelines in regards to a screening trial before proceeding with pump implantation2,3,7. Previously, trialing was considered to be the standard of care and critical in evaluation of a patient’s response and potential side effects from a certain agent. However, more recently, the PACC 2012 expert panel has called it debatable specifically in cancer-pain patients7. The issues of opioid-induced hyperalgesia (OIH) or disease progression is difficult to assess from a single-shot trial or even with a brief 72–96 hour infusion. A trial may also lead to underestimation of the failure rate and potential side effects. Still, the rationale for performing trials is based on the intuitive basis that it best mimics the system that would eventually be implanted7. Generally speaking, there is currently no solid base to either refute or adopt preimplantation trials but it is generally recommended although not required2. The British Pain Society expert panel suggested that trials should always be performed before the implantation of an intrathecal drug delivery system7.\n\nThere is a spectrum of intrathecal system options that range from a percutaneous catheter/external pump to a totally implanted system. Choice is dependent on a number of factors including, but not limited to, life expectancy, cost and availability of professional expertise and patient’s wishes and comfort level.\n\nIn general, percutaneous refers to a medical procedure in which a catheter or a medical device is introduced into the body via needle-puncture of the skin. This approach is favored because the catheter is easy to place (fluoroscopy can be used to confirm the position of the catheter and to maneuver it into the desired location) and is suitable for patients with limited life expectancy. The catheter is then attached to an external pump which delivers the medication directly into the intrathecal space. However, it is generally considered only a temporary option due to limitations such as frequent monitoring for infection, catheter migration, and patient’s immobility. The risks from this procedure are similar to performing a spinal anesthetic which include bleeding, infection, headache, spinal cord damage, etc.\n\nFully implanted drug delivery systems are suitable for long-term use. Mobility and functional activity are not particularly adversely affected by these systems. The implantation is performed by a skilled health care provider in an operating room under monitored anesthesia care with local anesthetic infiltration or general anesthesia. The patients require specialized care with a full multi professional infrastructure including regular follow-up appointments for pump refills. The pump can be programmed to be a fixed-rate, constant flow device or a variable rate pump with the option of boluses via patient controlled analgesia programming using a wireless radiofrequency transmitter given to the patient. Fixed rate delivery systems are less expensive than variable rate delivery systems but lack flexibility of medication delivery based on the need of a patient. These systems have a larger reservoir volume which allows for longer intervals between refills. In cases of suspected or actual medication overdose or implant malfunction, the pump’s drug reservoir has to be emptied and the device is interrogated by a designated health care professional.\n\n\nITDD devices\n\nThere are two systems available for implantable intrathecal devices–those that are programmable and those that are fixed-rate. In addition, some systems come with a bolus option that gives patients more control in managing their treatment. The needs of the patient as well as the suitability of the system for use with selected drugs need to be considered when deciding between the options.\n\nProgrammable devices such as Medtronic’s SynchroMed system deliver a specific amount of medication intermittently (Figure 2 and Figure 3). These systems are preferred by providers because they allow drug dosages to be changed without invasive intervention as the disease progresses and/or the patient builds tolerance to the medication. In cases of suspected or actual medication overdose or implant malfunction, the pump can be interrogated and deactivated without having to empty the drug reservoir. However, since it is battery driven, the life of this pump is usually 5–8 years and a surgical revision is required to implant a new pump. In addition, regular attendance for refilling is required because the drug reservoir tends to be smaller in volume than those found in fixed-rate devices.\n\nPump is programmable to deliver a specific amount of medication at different times and can be increased or decreased depending on the individual’s needs. Reprinted with the permission of Medtronic, Inc.\n\nThe pump has 20 and 40 cc reservoirs to extend the time between refills. Drug is refilled via the reservoir fill port. Reprinted with the permission of Medtronic, Inc.\n\nConstant flow systems such as Codman’s 3000 deliver a continuous amount of medication into the intrathecal space. These devices are favorable because they are generally less expensive, do not require a battery to operate and possess a larger drug reservoir that allows the drug to be delivered for a longer period of time before requiring a refill. However, constant rate devices are disadvantageous because they don’t give the flexibility to change the programming or a patient controlled bolus option for additional analgesia.\n\nThe Medtronic Synchromed II infusion system comes with a Patient Therapy Manager (PTM) which is a trademark of the Medtronic Corporation11. It is a remote hand-held device which communicates with the implanted pump and allows the patient to self administer boluses based on the medication limits which are pre-programmed by the clinical provider. This system gives patients the ability to activate a bolus option when adequate pain relief is not achieved. The clinical care provider has the ability to set the number of boluses a patient may receive in a 24 hour period. There is a lockout system in place with the programmer that does not allow the patient to self activate the bolus option more than what is allowed based on the settings created by the clinical provider. This allows for flexibility for the patients who have more intimate needs to control their pain based on certain body positions, movement, time of the day, and/or progression of their underlying disease. There are several safety options built in to the device such as the maximum number of boluses the patient may self administer in a 24 hour period, maximum dosage of the total drug in a 24 hour period, lockout in-between bolus activation, and the duration of time during which a bolus is administered.\n\n\nPlacement\n\nThe placement of ITDD consists of two parts: 1) placement of the catheter in the intrathecal space surrounding the spinal cord and 2) placement of the pump/reservoir in the abdomen/flank region. The pump is implanted in the lower abdomen and is attached to the intrathecal catheter which is tunneled under the skin from the patient’s back towards the abdomen. The procedure requires two surgical incisions: 1) Lower back where the catheter is inserted into the intrathecal space and is secured to the fascia. 2) The lower abdominal wall to create a pocket for the pump. The procedure generally takes 3 to 4 hours from start to finish12.\n\nPrior to the procedure, a thorough evaluation is performed by the anesthesiologist to ensure patient’s ability to undergo surgery and anesthesia. Physical exam, blood tests, electrocardiogram, and chest x-rays may be requested by the provider. Current systemic use of analgesics is discontinued as well as anticoagulants and antiplatelet therapies. The position of the pump is proposed and agreed upon by the patient and provider keeping the patients’ comfort, activity, need for future procedures and rehabilitation matters in consideration. The choice of medication is dependent on the success of past and present medication choices and acceptable side effects. This is usually pre-determined from either previous efficacy of a particular medication or during a trial period.\n\nImplantation of an ITDD system is a minor surgical procedure that is carried out in the operating room using aseptic precautions, including skin preparation, sterile draping, and the use of full surgical attire. The patient is positioned in a lateral decubitus position with the patient’s side for the pump pocket nondependent12. Real time fluoroscopy is used to identify the ideal location for placement of the catheter via a needle provided by the device manufacturer. A small skin incision is made in the middle of the back to expose the bony arch (lamina) of the vertebra and the catheter is placed in the subarachnoid, or intrathecal space and then anchored into the fascia to prevent migration.\n\nOnce the catheter is in place and secured, a tunneling device is used to pass the catheter under the skin from the spine to the abdomen where the pump will be implanted. A 4–6 inch skin incision is made in the side of the abdomen below the waistline. The surgeon then creates a pocket for the pump between the skin and muscle layers and the catheter is attached to the pump. Once the connector is attached to the pump, the pump is correctly positioned under the skin and secured to the thick fascia layer overlying the stomach muscles using suture loops. A space inside the pump called the reservoir is filled with the desired medication and the pump is programmed. The incisions in the back and abdomen are closed with sutures or staples and a dressing is applied. The patient is taken to the recovery area and is monitored per the standard requirements of the institution before being discharged.\n\nOften antibiotics are prescribed to prevent complications and a one week follow-up is recommended. There may be temporary restriction on daily activities to allow for healing of the wound and pump to settle. Once healed, no special care is necessary.\n\n\nDosing and titration\n\nAs mentioned previously, ITDD delivers medicine directly into the cerebrospinal fluid, bypassing the route that oral medication takes through the body. Generally 1/100th–300th of the amount of medication is used with the pump delivery as compared with the amount when taken orally. A survey conducted among a small group of physicians highly-experienced in ITDD revealed that morphine was the more frequently used and prescribed choice of medication, both alone and in combination with other drugs13. The concentrations and daily dosages administered of medications varied from patient to patient. Providers often adjusted the dosages of drugs or drug combinations by a fixed percentage, between 10% and 20%, whether increasing or decreasing the dosage. An average of five to eight adjustments in dosage was completed before a different drug or drug combination was introduced in cases of inadequate analgesia13.\n\n\nEfficacy\n\nStudies have suggested that ITDD is effective in providing pain relief to 60–80% of patients experiencing chronic malignant pain7. Long-term intrathecal application of opioid medications in cases of carcinoma pain is substantially more effective than its systemic application. Intrathecally applied opioids exert a strong analgesic effect via spinal receptors, without significantly influencing motor, sensory, and sympathetic reflexes. It has been shown to improve patients’ overall mood and reduce the incidence of systemic side effects. As a result, patients have reported an improvement in quality of life and the ability to participate more fully in daily activities.\n\n\nComplications\n\nAs with any surgical procedure, serious complications may arise post-implant that require immediate attention and perhaps the removal of the device.\n\nA retrospective study14 showed that the most common complication was patients’ adverse reaction to a drug. Pharmacological complications were the most common immediately post-implant and generally subsided as treatment continued. Serious complications include respiratory depression/arrest, anaphylaxis, and meningitis with introduction of contaminated solution. It was also found that certain opioids (the drugs of choice) were more prone to complications than others. For example, the incidence of tip catheter granulomas is higher with morphine and hydromorphone as compared to fentanyl which has been upgraded to a first-line drug in the algorithm due to that specific advantage7. Table 5 below lists the most common drug related adverse reactions and it is compiled based on the data presented in the PACC 2012.\n\nCatheter and pump related malfunctions are one of the main sources of complications. In fact, one study14 found catheter related complications as the most common cause of repeat surgery. Despite suturing the catheter to the underlying fascia, coiling of the tube and leakage at the connection site remains a problem. A study reviewing complications from long-term intrathecal drug therapy found that the annual rate for complications requiring surgical procedure was 10.5%, with 35% being pump related and 65% catheter related2. Errors in pump programming have also been reported and resulted in incorrect flow rates. Device related complications have led to infections (meningitis), pocket abscesses, bleeding, pain and discomfort, and blood or fluid in the pump’s pocket. Table 6 lists the most common device related side effects.\n\nAlthough rare, patient induced complications generally involve infections at implant location. Immune cells form a mass, known as a granuloma, around the catheter tip to wall off foreign substances. However, the incidence of infections in all cases reviewed during a study was 0.7% per year15, with all infections appearing within the first three months post-implant. Factors that increased the potential risk of complication included psychological problems, obstructive sleep apnea, immunosuppression, smoking, diabetes active infections, bleeding disorders, and concurrent anticoagulation therapy2,7.\n\nThe FDA and the United States Pharmacopoeia have issued standards on compounded sterile products that have clinical, legal, and practical significance7,9. These standards apply to compounding of solutions by various routes, including intrathecal administration. There are strict provisions in place for considerations such as: training of personnel, a segregated sterile compounding area, air quality of the compounding area, certification and calibration of equipment, a cleaning and disinfection program, and a quality assurance program7. Any violation of any of the provisions can result in a civil and criminal law suit in addition to the suspension of the privileges to dispense compounded medications. An outbreak of fungal meningitis resulted in 2012 when several lots of contaminated methylprednisolone vials were used for epidural steroid injections in various states which were obtained from a compounding pharmacy. Shortly following the outbreak, the device manufacturer Medtronic released an “Urgent Medical Device Safety Notification” in November 2012 discouraging the use of unapproved drugs with its delivery system9.\n\n\nCost effectiveness\n\nWith estimates of pain care treatment costs exceeding $1 billion annually in the United States, cost and efficacy are important factors in clinical decision making. Although the initial cost of intrathecal drug delivery is substantially more, maintenance costs over time are significantly lower than conventional routes of administration. Cost analyses by the group concluded that intrathecal delivery is the most cost-effective route of opioid administration for patients who require long-term management of cancer (≥ 3–6 months) or nonmalignant pain (≥ 11–22 months)16.\n\nAn analysis by Kumar et al. showed that intrathecal therapy is a cost-effective method of treating chronic nonmalignant pain caused by failed back syndrome and that the break-even point occurs at 28 months, after which conventional pain therapy is the more expensive treatment option17. A separate analysis by Scott Guillemette from Ingenix Consulting on patients suffering from failed back surgery syndrome concluded with similar results. Despite the higher upfront costs, patients utilizing intrathecal drug delivery returned to ‘normal living’ more quickly than those with conventional therapy and the break-even point between the two points occurred between months 19 and 2018.\n\n\nConclusion\n\nThe treatment of pain poses an ongoing challenge to the healthcare field and better prevention, assessment and treatment of pain is needed. Intrathecal drug delivery can become a critical component of this transformation in treatment when more conservative forms of pain management have failed to provide adequate pain relief. In comparison to more conservative delivery methods such as drugs administered orally or intravenously, ITDD is more effective because the medication is directly introduced into the subarachnoid space8. The spinal cord is a key anatomical site for pain processing and more specifically the dorsal horn of the spinal cord within the intrathecal space5,8. The advantage of delivering the medication directly into the CSF in smaller dosages is that it may reduce the systemic side effects and the direct delivery to the site of action allows a more rapid and effective response. As a popular alternative for patients experiencing cancer or chronic pain, intrathecal analgesia has become a therapeutic option for patients who have exhausted all other treatment avenues as well as patients experiencing side effects from their current treatment options. However, there are limitations such as adequate patient selection, ability to tolerate the procedure, and costs. There are also potential complications related to the device, drug, and procedure that may require hospitalization and immediate removal/revision of the device. Despite numerous studies involving existing and novel drugs, only limited data is available to address the issue of medication safety, efficacy, stability and compatibility with the intrathecal drug delivery system. As this option becomes more popular with patients and physicians, future research should focus on cross-site studies that would provide a more accurate outlook on this type of delivery.", "appendix": "Author contributions\n\n\n\nGuarav Bhatia edited the manuscript to its present state. Mary E Lau formulated the initial draft of the article. Padma Gulur conceptualized the article and edited the manuscript to its present state. All authors agreed to the publication of 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 grants were involved in supporting this work.\n\n\nReferences\n\nCohen SP, Dragovich A: Intrathecal analgesia. Anesthesiol Clin. 2007; 25(4): 863–882. PubMed Abstract \| Publisher Full Text\n\nStearns L, Boortz-Marx R, Du Pen S, et al.: Intrathecal drug delivery for the management of cancer pain: a multidisciplinary consensus of best clinical practices. J Support Oncol. 2005; 3(6): 399–408. PubMed Abstract\n\nSmith HS, Deer TR, Staats PS, et al.: Intrathecal Drug Delivery. Pain Physician. 2008; 11(2 Suppl): S89–S104. PubMed Abstract\n\nvan den Beuken-van Everdingen MH, de Rijke JM, Kessels AG, et al.: Prevalence of pain in patients with cancer: a systematic review of the past 40 years. Ann Oncol. 2007; 18(9): 1437–49. PubMed Abstract \| Publisher Full Text\n\nVranken JH, van der Vegt MH, Kal JE, et al.: Treatment of neuropathic cancer pain with continuous intrathecal administration of S +-ketamine. Acta Anaesthesiol Scand. 2004; 48(2): 249–252. PubMed Abstract \| Publisher Full Text\n\nInternational Anesthesia Research Society. Global year against cancer pain: epidemiology of cancer pain. Refernce Source\n\nDeer TR, Prager J, Levy R, et al.: Polyanalgesic Consensus Conference 2012: Recommendations for the management of pain by intrathecal (intraspinal) drug delivery: report of an interdisciplinary expert panel. Neuromodulation. 2012; 15(5): 436–64; discussion 464–6 PubMed Abstract \| Publisher Full Text\n\nAhmed S: Chapter 13: Intrathecal drug delivery for chronic pain management. Translational Pain Research 2006; 1. : Nova Science Publishers, Inc.\n\nMedtronic. Urgent medical device safety notification: use of unapproved drugs with the SynchroMed Implantable Infusion Pump. Reference Source\n\nWinkelmuller M, Winkelmuller W: Long-term effects of continuous intrathecal opioid treatment in chronic pain of nonmalignant etiology. J Neurosurg. 1996; 85(3): 458–67. PubMed Abstract \| Publisher Full Text\n\nMedtronic. Personal therapy manager for SynchroMed II physician manual. Reference Source\n\nRathmell JP: Atlas of image-guided intervention in regional anesthesia and medicine. Second Edition. Lippincott Williams & Wilkins. Philadelphia, PA. 2012. Reference Source\n\nHassenbusch SJ, Portenoy RK: Current practices in intraspinal therapy – A survey of clinical trends and decision making. J pain Symptom Manage. 2000; 20(2): S4–S11. PubMed Abstract\n\nKrakovsky AA: Complications associated with intrathecal pump drug delivery: a retrospective evaluation. Am J of Pain Med. 2007; 17(1): 4–10.\n\nFluckiger B, Knecht H, Grossmann S, et al.: Device-related complications of long-term intrathecal drug therapy via implanted pumps. Spinal Cord. 2008; 46(9): 639–43. PubMed Abstract \| Publisher Full Text\n\nMueller-Schwefe G, Hassenbusch SJ, Reig E: Cost effectiveness of intrathecal therapy for pain. Neuromodulation. 1999; 2(2): 77–87. PubMed Abstract \| Publisher Full Text\n\nKumar K, Hunter G, Demeria DD: Treatment of chronic pain by using intrathecal drug therapy compared with conventional pain therapies: a cost-effectiveness analysis. J Neurosurg. 2002; 97(4): 803–810. PubMed Abstract \| Publisher Full Text\n\nGuillemette S: American Academy of Pain Medicine (AAPM) 26th Annual Meeting. February 3–6, San Antonio, Texas 2010. Reference Source" }	[ { "id": "885", "date": "10 Apr 2013", "name": "Michael Erdek", "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\nPage 3: There can be a need for increasing doses of opioids as tolerance develops.Page 5: There is a difference in thickness of the 20cc and 40cc pumps.Page 6: The lamina is not routinely exposed unless laminotomy leads are being placed.Page 7: Reference Smith TJ et al. (2002) J Clin Oncol. for efficacy.Page 7: Reference Follett K et al. (2004) Anesthesiology for complications.Page 7: (Table 5) – Ziconotide is not an opioid.", "responses": [ { "c_id": "482", "date": "10 Jun 2013", "name": "Padma Gulur", "role": "Author Response", "response": "Thank you very much for your response. Page 3: There can be a need for increasing doses of opioids as tolerance develops. We understand and agree with your comment; however, we are unsure as to where you feel this statement should to be added within the article. Page 5: There is a difference in thickness of the 20cc and 40cc pumps. We understand and agree with your comment; however, we are unsure as to where you feel this statement fits in the article.Page 6: The lamina is not routinely exposed unless laminotomy leads are being placed.6.2 Operating room procedure - We have changed the sentence to say that the incision is made to expose the fascia over the bony arch (lamina).Page 7: Reference Smith TJ et al. (2002) J Clin Oncol. for efficacy. A discussion of this article has been added into the efficacy section to show that ITDD is a better system of pain management than comprehensive medical management for patients with cancer pain. Page 7: Reference Follett K et al. (2004) Anesthesiology for complications. After reviewing this article, it has been added into our article that in order to further decrease the chances of infection, there must be an adherence to guidelines and recommendations regarding surgical site infections. Page 7: (Table 5) – Ziconotide is not an opioid. Ziconotide has been taken out of the chart." } ] }, { "id": "905", "date": "22 Apr 2013", "name": "Arun Bhaskar", "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 gives a broad overview of the various types of ITDD systems used in the management of complex cancer pain. It also discusses the advantages, the common side-effects and complications, the technique and cost-effectiveness of using ITDD systems. Overall the article is useful in providing an understanding for doctors and nurses not familiar with the intrathecal drug delivery systems. The article is written at a basic level and although it is useful to beginners and people who are not well versed in ITDD, it does not deliver much to experienced practitioners. I have reservations as I feel that this review article would have been of a higher standard if it was more comprehensive.", "responses": [ { "c_id": "481", "date": "10 Jun 2013", "name": "Padma Gulur", "role": "Author Response", "response": "Thank you very much for your response. This article was written for general practitioners to gain an understanding of ITDD systems for cancer pain." } ] } ]	1	https://f1000research.com/articles/2-96
https://f1000research.com/articles/3-173/v1	25 Jul 14	{ "type": "Software Tool Article", "title": "Explain your data by Concept Profile Analysis Web Services", "authors": [ "Kristina Hettne", "Reinout van Schouwen", "Eleni Mina", "Eelke van der Horst", "Mark Thompson", "Rajaram Kaliyaperumal", "Barend Mons", "Erik van Mulligen", "Jan A. Kors", "Marco Roos", "Reinout van Schouwen", "Eleni Mina", "Eelke van der Horst", "Mark Thompson", "Rajaram Kaliyaperumal", "Barend Mons", "Erik van Mulligen", "Jan A. Kors", "Marco Roos" ], "abstract": "The Concept Profile Analysis technology (overlapping co-occurring concept sets based on knowledge contained in biomedical abstracts) has led to new biomedical discoveries, and users have been able to interact with concept profiles through the interactive tool “Anni” (http://biosemantics.org/anni). However, Anni provides no way for users to save their procedures, results, or related provenance. Here we present a new suite of Web Service operations that allows bioinformaticians to design and execute their own Concept Profile Analysis workflow, possibly as part of a larger bioinformatics analysis. The source code can be downloaded from ZENODO at http://www.dx.doi.org/10.5281/zenodo.10963.", "keywords": [ "Concept Profile Analysis (CPA) has proven a powerful tool for interpreting and prioritizing results of bioinformatics analysis", "and for linking data sets based on the best “educated guess” when precise links are not available. The technology uses the vector space model to relate concepts (such as genes and biological processes) mined from the literature to each other. Vectors can be compared efficiently and transparently1", "and the model yields a measure of the strength of the relationship between concepts. We call these vectors “concept profiles”. The CPA algorithms have for example been successfully applied to compare microarray studies2", "for predicting proteins putatively associated with muscular dystrophy pathways3", "and for associating chemical structures with gene expression data4." ], "content": "Introduction\n\nConcept Profile Analysis (CPA) has proven a powerful tool for interpreting and prioritizing results of bioinformatics analysis, and for linking data sets based on the best “educated guess” when precise links are not available. The technology uses the vector space model to relate concepts (such as genes and biological processes) mined from the literature to each other. Vectors can be compared efficiently and transparently1, and the model yields a measure of the strength of the relationship between concepts. We call these vectors “concept profiles”. The CPA algorithms have for example been successfully applied to compare microarray studies2, for predicting proteins putatively associated with muscular dystrophy pathways3, and for associating chemical structures with gene expression data4.\n\nThe standalone application Anni10 supports a number of standard CPA operations. For example, to perform pathway analysis for a gene expression experiment, a user first provides a list of gene database identifiers for the most significantly expressed genes. Anni uses these identifiers to query the concept profile database for the corresponding concept profiles, and subsequently constructs a “concept set” of these profiles. To match the list of genes with pathways, the user performs the operation “match concept sets” for the gene concept set with a predefined concept set of the category “Gene Ontology (GO) biological process”. Note that we refer here to GO concept profiles. The concept profile matching scores between the two concept sets are calculated by Anni, resulting in a ranked list of GO biological processes for the gene list. Finally, literature evidence in the form of documents containing co-mentions of the gene and biological processes can be retrieved by Anni from a supporting documents database, or from documents providing enough statistical evidence to support the gene-biological process associations without actually mentioning the gene and the biological process together in an abstract.\n\nHere we present a new suite of Web Service (WS) operations that allows bioinformaticians to design and execute their own CPA workflow outside the Anni Web tool, possibly as part of a larger bioinformatics analysis. The WS was designed according to the outcome of an Anni usage analysis, where the common user and machine operations were identified.\n\nWe implemented the CPA WS using Java, Model-View-Controller (MVC) Spring framework, and Apache Tomcat following the Java API for XML WS (JAX-WS) specifications. We compiled the Anni Java code for the different operations into separate libraries, for which wrappers were written in Java. Spring MVC was used as a WS interface to remote applications. The WS was implemented according to the JAX-WS standard, enabling an auto-generated WSDL specification and use of Java Annotations to specify operations. Apache Tomcat was used for deployment. The CPA WS uses a database of indexed PubMed records. The thesaurus behind the Anni Web application was converted to Simple Knowledge Organization System (SKOS), and the SKOS concept IDs were implemented as resolvable Unique Resource Identifiers leading to a Virtuoso Universal Server triple store.\n\nAs an example on how to work with the CPA WS we implemented several workflows in the workflow management system Taverna workbench v 2.45 following the best practices for workflow design6. The whole suit of CPA workflows consists of 11 workflows collected in a myExperiment pack [http://www.myexperiment.org/packs/368]. These workflows are of two different types: 1) nine workflows calling one WS operation, and 2) two pipelines of nested workflows calling more than one WS operation. The workflows of type 1 are the building blocks to make pipelines of type 2, and were implemented with re-usability in mind.\n\nHere we describe the workflow “Match concept profiles with predefined set” (Figure 1) in order to illustrate the design and use of the WS and workflows. The workflow invokes the WS operation “getSimilarConceptProfilesPredefined”. The operation takes three input parameters, which can be accessed using the XML splitter function in Taverna. The user specifies the concept(s) to be matched (“Query concept IDs”), the concept set to match against (“Match concept set”), and a cutoff number of matched concepts to return (“Cutoff”).\n\nBlue boxes represent the workflow inputs and outputs, green box the WS invocation, and purple boxes the XML splitters for the inputs and outputs of the WS operation. The workflow is available at http://www.myexperiment.org/workflows/3396.\n\nOpening the “Run workflow” window in Taverna will result in showing the structured annotations for the whole workflow and the input parameters, as well as the example values (Figure 2). WS functional annotations can be accessed via the “Details” tab in Taverna (Figure 3). When the workflow is run, it will produce a ranked list of concepts associated to the query concept(s), and their similarity scores.\n\nDetailed, structured descriptions for the whole workflow and its input parameters, with example values are shown in the window.\n\nA detailed description of the function of the WS operation is shown in the window.\n\nThe above described workflow executes the core functionality of concept profile matching. The other WS operations implement functionality such as explaining the association found (by listing the common concepts contributing most to the score) and showing the literature evidence (by retrieving the links to the abstracts in PubMed). Workflows implementing these WS can be coupled to the “Match concept profiles with predefined set” workflow to form a pipeline of nested workflows. Examples of such pipelines are the “GWAS to biomedical concept” nested workflow, which performs Single Nucleotide Polymorphism annotation (SNP), and the “Annotate gene list with top ranking concepts” nested workflow for gene annotation (Figure 4).\n\nBlue boxes represent input and output parameters, purple boxes the local Taverna worker services, yellow boxes the Xpath services for fast XML parsing, and grey boxes the constant values. The workflow is available at http://www.myexperiment.org/workflows/3921.\n\n\nDiscussion\n\nThe CPA WS and workflows raise the level of reproducibility of bioinformatics experiments that make use of CPA compared to Anni, and the CPA WS can more easily be used together with other tools. For example, CPA-based SNP annotation can be performed with the CPA WS by coupling an external WS to map the SNP identifiers to Entrez gene identifiers7. With Anni, the SNP to Entrez gene identifier analysis would have to be performed separately, decreasing the reproducibility.\n\nSome of the functionalities in Anni have not been migrated to the WS. For example, Anni provides a function for hierarchical clustering of the results. Clustering is not a CPA function by itself, but we are considering to implement workflows that perform this function. We are also working on a workflow implementation of the process that creates the data underlying the Anni WS, possibly using the recently developed text-mining workbench Argo8, allowing for more flexibility in performing CPA9. Specialization of the underlying resources for services to use in specific research domains, such as plant breeding or metabolomics, is a topic for future work.\n\n\nConclusions\n\nBy creating a WS building upon the Anni interactive tool, we made available the CPA technology in a way that users can easier integrate the technology with other software and save their procedures, results and related provenance.\n\n\nSoftware availability\n\nCPA WS: https://www.biocatalogue.org/services/3559\n\nhttps://trac.nbic.nl/conceptprofileminingws/\n\nWorkflows: http://www.myexperiment.org/packs/368\n\nhttp://www.dx.doi.org/10.5281/zenodo.1096311\n\nApache 2.0", "appendix": "Author contributions\n\n\n\nKH designed the workflows, performed the usage analysis for the WS, and wrote the manuscript. RS implemented the WS. EM helped in the design of the workflows, the usage analysis and testing of the WS. EH converted the thesaurus into SKOS format and set up the concept triple store. MT and RK helped in implementing the WS and setting up the concept triple store. BM, EvM, and JK helped in the usage analysis and testing of the WS. MR conceived the study and helped design the workflows. All authors approved the final version 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 work in this paper was funded by the Seventh Framework Programme of the European Commission (Digital Libraries and Digital Preservation area ICT-2009.4.1 project reference 270192) (Wf4Ever), and grant agreement No. 305444 (RD-Connect).\n\n\nAcknowledgements\n\nWe would like to thank Peter-Bram ’t Hoen for his comments about the WS design and functionality.\n\n\nReferences\n\nJelier R, Schuemie MJ, Roes PJ, et al.: Literature-based concept profiles for gene annotation: the issue of weighting. Int J Med Inform. 2008; 77(5): 354–362. PubMed Abstract \| Publisher Full Text\n\nJelier R, ’t Hoen PA, Sterrenburg E, et al.: Literature-aided meta-analysis of microarray data: a compendium study on muscle development and disease. BMC Bioinformatics. 2008; 9: 291. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Haagen HH, ’t Hoen PA, de Morrée A, et al.: In silico discovery and experimental validation of new protein-protein interactions. Proteomics. 2011; 11(5): 843–853. PubMed Abstract \| Publisher Full Text\n\nHettne KM, Boorsma A, van Dartel DA, et al.: Next-generation text-mining mediated generation of chemical response-specific gene sets for interpretation of gene expression data. BMC Med Genomics. 2013; 6(1): 2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolstencroft K, Haines R, Fellows D, et al.: The Taverna workflow suite: designing and executing workflows of Web Services on the desktop, web or in the cloud. Nucleic Acids Res. 2013; 41(Web Server issue): W557–W561. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHettne KM, Wolstencroft K, Belhajjame K, et al.: Best Practices for Workflow Design: How to Prevent Workflow Decay. In Proceedings of SWAT4LS 2012, 2012. Reference Source\n\nHettne KM, Dharuri H, van Schouwen R, et al.: Explaining genome-wide association study results using concept profile analysis and the Kyoto Encyclopedia of Genes and Genomes pathway database. In Proceedings of BioLINK SIG 2013, 2013; page 60. Reference Source\n\nRak R, Batista-Navarro RT, Carter J, et al.: Processing biological literature with customizable web services supporting interoperable formats. Database(oxford). 2014; 2014: pii: bau064. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan der Horst E, Roos M, Hettne K: Workflows and services for concept profile generation. F1000Posters. 2014; 5(33). Reference Source\n\nJelier R, Schuemie MJ, Veldhoven A, et al.: Anni 2.0: a multipurpose text-mining tool for the life sciences. Genome Biol. 2008; 9(6): R96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHettne KM, van Schouwen R, Mina E, et al.: New suite of Concept Profile Analysis Web Services. ZENODO. 2014. Data Source" }	[ { "id": "5585", "date": "05 Aug 2014", "name": "Karin Verspoor", "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 describes a suite of services that bring the previously existing concept profile analysis tool \"Anni\" to the web. These web services enable a range of bioinformatics analysis tasks, and can be embedded in other workflows. The authors provide examples of implementing workflows within the Taverna workbench which nicely illustrate the applicability of the services.The authors provide links to both the software and the Taverna workflows used as examples, which is very good.The article could be improved by providing a more clear distinction between what can be achieved with the web services, and how the operations are implemented. The very first paragraph of the article, for instance, provides a strongly methodological description of the tool (introducing the vector space model, and the notion of comparing vectors). A more application-oriented oriented motivation for the tool would be preferable right at the beginning; the details of e.g. how \"pathway analysis\" is supported (via the concept profiles) can be provided after the general rationale for the tool. Note that the authors also mention a use case for \"linking data sets\" but do not appear to provide a concrete example of where this might be relevant. Similarly, the technical specifications mix somewhat user and developer functionality; some reorganisation could improve the manuscript. Where they mention \"the user\", is this a 'real' user (i.e. someone interested in bioinformatics analysis), or a computational user/developer?Where the authors state, \"The concept profile matching scores between the two concept sets\", does this refer to a pairwise comparison between the elements of the two sets, or a more holistic/aggregated comparison across the complete sets? Please clarify.Figure 2 might be more interesting if it showed an example of a \"Match concept set\" rather than the \"Cutoff\" value. How are such concept sets defined? On the other hand, parameters such as \"cutoff\" should be explained, at least in the figure caption, for completeness.There are several details provided that are not fully explained. Specifically, there is a mention of \"the XML splitter function in Taverna\" for input parameters. Again, please consider whether this is important for a user to understand, or rather for a developer to understand the code. Also, what are \"WS functional annotations\"? (cf. also the reuse of the term \"annotation\" with possibly a different sense in \"Single Nucleotide Polymorphism annotation\" and \"gene annotation\"). What do you refer to with \"the data underlying the Anni WS\"?The conclusions are justified, if a bit unspecific. A more targeted summary of the functionality would be preferable, rather than \"we made available the CPA technology\".Note a few small English usage issues. (1) \"considering to implement\" should be \"considering implementing\" (2) \"users can easier integrate\" should be \"users can more easily integrate\".", "responses": [] }, { "id": "5805", "date": "18 Sep 2014", "name": "Naoaki Okazaki", "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 presents a suite of Web services with which users can design and execute their own workflow for Concept Profile Analysis. The paper includes examples of running the Web services on Taverna workbench. The motivation of promoting the Web services presented in this paper is unclear to me. The abstract says, “However, Anni provides no way for users to save their procedures, results, or related provenance.” I guess that the authors wrote this sentence in order to explain the advantage of using the Web services over Anni. However, I am not sure whether bioinformaticians prefer the Web services to Anni in their analyses. Some may find Anni more useful than the Web services because Anni was designed to perform pre-defined workflows. For this reason, I am also not sure whether the conclusions are justified sufficiently on the basis of the developed Web services. The impact of this paper would be greater if the authors could explain the sales point(s) of this software (especially to bioinformaticians, real users) more concretely. The explanation of how the Web services work on Taverna workbench is helpful for us to understand what can be achieved by the Web services. However, this paper lacks the comprehensive detail of the Web services, e.g., what kinds of services were implemented, what is the functionality of each service, what is the underlying research technologies for implementing the each service (although software technologies such as Apache Tomcat are explained). These details may be useful for readers to imagine the possible way to integrate the Web services with other tools. Minor comment: It would be better if this paper could introduce Concept Profile Analysis (CPA) with references.", "responses": [] } ]	1	https://f1000research.com/articles/3-173
https://f1000research.com/articles/3-171/v1	25 Jul 14	{ "type": "Case Report", "title": "Case Report: Reversible cabergoline-associated cardiac valvulopathy post drug discontinuation", "authors": [ "Chris G. Yedinak", "Shirley McCartney", "Troy H. Dillard", "Kevin S. Wei", "Maria Fleseriu", "Chris G. Yedinak", "Shirley McCartney", "Troy H. Dillard", "Kevin S. Wei" ], "abstract": "We present a case of a 21 year old male patient diagnosed with a 2.2 cm prolactin-secreting adenoma in contact with the optic chiasm. The patient was treated with up to 6mg/week of cabergoline (total cumulative dose 814 mg) and developed mild valvulopathy. Valvulopathy was subsequently reversed after discontinuation of cabergoline therapy.", "keywords": [ "prolactin-secreting", "adenoma", "valvoplasty" ], "content": "Presentation\n\nA 21 year old male patient with a history of delayed puberty and a 2.2 cm pituitary adenoma was referred for evaluation. At presentation, prolactin level was found to be greater than 1000 ng/ml (normal range: 3–13 ng/ml) on diluted testing. Treatment with the dopamine agonist (DA) cabergoline was initiated and over a 10-month period, the macroadenoma shrank substantially to < 1cm (as observed by magnetic resonance imaging). Total cabergoline dose at that time was 4.5 mg/week, in three equally divided weekly doses. Neither prolactin level nor testosterone level normalized. The patient reported no symptoms as would have been expected of a prolactinoma of this size; he denied visual field changes or deficits, galactorrhea, breast discomfort, fatigue, temperature dysregulation, weight changes or fluctuations, hair loss, weakness, memory changes or muscle loss. He felt his libido was somewhat diminished and morning erections were absent. He did report headaches at presentation that resolved after 10 months of treatment with cabergoline.\n\nMedical history was significant only for reactive airway disease for which he was treated with albuterol (as needed), which was seldom used. He was not sexually active, was a non-smoker with infrequent alcohol use and no caffeine intake. The patient confirmed a family history of cancer and arthritis in a grandmother and hypertension in a grandfather, but a negative family history of cardiac disorders, pituitary masses or known heritable diseases.\n\nOn physical exam he was normotensive 119/82 mmHg, pulse 81, weight 90.7 Kg, height 1.78 m (body surface area 2.06 m2). He did not have visual field deficit to confrontation and had a normal cardiac exam without murmur, rub or gallop. Chest was clear to auscultation and breast exam revealed no gynecomastia, tenderness or nipple discharge. He was void of body hair on his chin, chest, arms and legs. The genitourinary exam revealed pre-pubertal testes with no pubic hair and small penile size. The remainder of his physical exam was unremarkable.\n\nLaboratory evaluation indicated a testosterone level 51 ng/dL (normal range: 241–950 ng/dL) with follicle-stimulating hormone (FSH) and luteinizing hormone (LH) of 1 mlU/mL (inappropriately low; normal adult male range FSH 1.5–12.4 mIU/ml and LH 2.1–4.7 mIU/ml). Thyroid axis was noted to be intact with thyroid stimulating hormone of 2.1 µlU/ml (normal range: 0.34–5.6 µlU/ml) and free thyroxine (T4) of 0.7 ng/dL (0.6–1.2 ng/dL). Prolactin level was 55 ng/ml (normal range: 3–13 ng/ml). A cosyntropin stimulation test was performed to evaluate the hypothalamic-pituitary-adrenal (HPA) axis function with a normal baseline cortisol level at 9:20 am of 14.3 µg/dL and adrenocorticotropin hormone (ACTH) of 41 pg/mL (< 46 pg/ml). His cortisol level was minimally blunted at 17.2 µg/dL (normal range: > 18 µg/dL) post stimulation with low dose (1 μg) Cortrosyn. IGF-1 was low at 84 ng/mL (age and gender adjusted normal 116–358 ng/mL). Magnetic resonance (MR) imaging prior to cabergoline treatment, indicated a 2.2 cm tumor in contact with optic chiasm (Figure 1A, B); follow up MR imaging after 10 months of cabergoline treatment revealed residual tumor on the left side of the gland (< 1cm) without optic chiasm involvement (Figure 1C, D). The patient’s bone age was evaluated as 16.5 years.\n\nMagnetic resonance imaging: pre-treatment (A) sagittal T1 and (B) coronal T1, and 2 years post-treatment (C) sagittal T1 and (D) coronal T1.\n\n\nDiagnosis\n\nThe presence of a pituitary macroadenoma, hypogonadism and a single measurement of serum prolactin > 250 ng/mL confirmed a diagnosis of prolactinoma1. Likewise, drug-induced hyperprolactinemia, elevation related to stalk effect and drug effects were excluded by this degree of elevation.\n\nThere was no evidence of co-secretion with growth hormone and the primary diagnosis remained hyperprolactinemia related to a prolactin-secreting pituitary adenoma with central hypogonadotrophic hypogonadism. Given the significant tumor shrinkage, DA therapy with cabergoline was continued without dose change. No testosterone replacement was initiated, in anticipation of further tumor response to DA therapy and subsequently normalization of both the prolactin level and spontaneous puberty. However, low bone age raised concern for growth hormone deficiency versus hypogonadism from possible pituitary damage and low dose depot testosterone (50 mg intramuscular monthly) and growth hormone replacement (0.2 mg/daily) were started for a trial of 6 months. As prolactin level remained elevated after a further 3 months of treatment, cabergoline dose was increased to 5 mg in a divided dose twice a week. He continued to tolerate DA therapy without side effects.\n\nAt follow up testing, after an additional 3 months of treatment, HPA axis function had not improved and he was treated with low dose glucocorticoids (hydrocortisone 10 mg daily) and thyroid (levothyroxine 50 mcg daily) replacement. After a further 12 months of treatment, prolactin level was reduced to 22 ng/ml (normal range: 3–17 ng/ml), but did not normalize; however tumor remained stable. Cabergoline dose was increased to a total of 6 mg weekly in three divided doses. After a total of 2 years of treatment with both DA and testosterone replacement, he began to gain muscle mass and achieved Tanner 4/5 with some penile growth.\n\nAfter an accumulative dose of 814 mg cabergoline over 4 years of treatment, the patient was found to have developed a faint systolic murmur on auscultation. A 2D, color and spectral Doppler echocardiogram indicated mild non-coaptation of the mitral leaflets associated with slight apically displacement in systole, normal valve thickness and excursion resulting in mild mitral regurgitation (Figure 2A, B). Ejection fraction was estimated at 60–65% with normal left ventricular function. Doppler measures of mitral valve peak E 1.08 m/s and peak A 0.29 m/s with E/A ratio 3.75 where E (early wave) represents passive filling of the ventricle and A (atrial wave) the active filling with atrial systole. Classically, the E-wave velocity is slightly greater than that of the A wave. All other values demonstrated normal structure and function. Changes were reported to be consistent with cabergoline therapy. Cabergoline was subsequently discontinued and DA therapy was continued using bromocriptine 5 mg/daily. The patient’s prolactin level increased to 70 ng/ml (normal range: 3–13 ng/ml) 2 months after switching from cabergoline therapy to bromocriptine.\n\nEchocardiograms: baseline (A and B) demonstrates apical displacement during systole with mild non-coaptation of the mitral leaflets and mild mitral regurgitation (arrows). Follow-up (C and D), 8 months after discontinuing cabergoline and starting bromocriptine, demonstrates a normal valve without mitral regurgitation (arrows).\n\nMR imaging performed after 6 months of treatment with bromocriptine was unchanged. All other pituitary hormonal levels were stable and no other concomitant medications had been used. Repeat echocardiography (performed in the same modes by the same sonographer and reviewed by the same cardiologist), revealed normal mitral valve function with Doppler measures of mitral valve peak E of 0.68 m/s and peak A 0.32 m/s and E/A ratio 2.14 but moderately reduced global RV systolic function and LV function of 55–60% (Figure 2C, D). Clinical exam, including vital signs, was unchanged between these two visits, the bromocriptine dose was increased to 10 mg daily and the patient was referred for cardiac consultation.\n\nAt cardiac consultation (3 months post bromocriptine dose increase) the patient did not have a murmur on exam. On review of the echocardiogram no appreciable evidence was found for ongoing disease and follow up echocardiography was recommended after a further 6 months. At this exam, a normal LV systolic function was found with an LV ejection fraction estimated at 62% using biplane Simpson’s method. No other echocardiographic abnormalities were found and improvement from previous echocardiograms was reported.\n\n\nDiscussion\n\nCabergoline has been implicated in the induction of fibrotic cardiac valvulopathy when used in high doses mostly for the treatment of Parkinson’s disease2,3 but also in treatment of prolactinomas4–6. Other studies have not shown a significant increase in cardiac valvular disease when cabergoline was used in patients with pituitary diseases7–10. The pathophysiologic mechanism of DA-related valvular heart disease is thought to be related to interactions of the drug with serotonin (5-HT) receptors, particularly 5-HT2B receptors. This receptor is present in fibroblasts on heart valves in high concentrations, but is also present in pulmonary arteries11–13.\n\nTwo independent researchers in large European studies14,15 reported an association between high doses of DAs (associated with the treatment of Parkinson’s disease) with potent 5-HT2B agonist activity and cardiac valve disease: particularly of the mitral, aortic and tricuspid valves. Activation of these receptors has been demonstrated to lead to excess cell division and overgrowth valvulopathy and dysfunction16. Only pergolide16 and cabergoline demonstrate 5-HT2B receptor activity. In a study by Zanettini et al. rates of drug induced valvulopathy occurred in 28.6% of treated patients14.\n\nBromocriptine was thought to be devoid of this activity15, but partial 5-HT2B receptor activity was subsequently demonstrated in porcine models17. However, studies of this effect for bromocriptine in patients treated for prolactinomas are scarce. Two studies of subjects treated long term (mean 54.8 vs 58.98 months) are available: 55 subjects (58.98 months) that reported significantly higher end diastolic intraventricular septal thickness with bromocriptine5; and 19 cases (54.8 months) with a higher prevalence of tricuspid regurgitation. Several case reports of bromocriptine use for Parkinson’s disease reported a higher incidence of pleuropericarditis in association with long term use5,18.\n\n\nManagement\n\nAlthough there are reports of reversibility of valvulopathy with other DAs19, this is the first case of reversal of valvular abnormalities after stopping cabergoline treatment in a patient with a prolactin-secreting adenoma. It is unclear if the risk of cardiac valvulopathy is associated with a high cumulative dose19 or is gender related20. However, this case highlights the potential reversibility of mild valvulopathy associated with cabergoline therapy if treatment is discontinued prior to the onset of severe structural abnormalities.\n\nReversibility of ergot-derived DA and 5HT2B agonist-induced valvular heart disease is infrequently documented and mostly limited to patients with Parkinson’s disease or weight loss treatment. In a rodent model, 12 weeks of serotonin injections induced both aortic and mitral regurgitation. Eight weeks after cessation of serotonin injections, the prevalence of valvulopathy was no longer higher than control and valvular thickness returned to baseline19. In 50 patients previously given fenfluramine or dexfenfluramine, valvular regurgitation in 17 of the 38 (45%) patients with mitral regurgitation and 19 of the 43 (44%) patients with aortic regurgitation improved after stopping the drug12. Regression of mitral valve disease was also noted in 4/10 patients who had discontinued pergolide (for Parkinson’s) in the previous 4–6 months21. To the best of our knowledge, only one case (1 of 4 treated for Parkinson’s disease) of reversible mitral valve disease related to cabergoline discontinuation has been described2.\n\nThe potential for cardiac valve effects in patients treated for prolactinomas has been previously reported3,4,22, but subsequent studies have not confirmed these findings7,9,23. Cabergoline remains an effective treatment to normalize prolactin levels and for tumor shrinkage in patients with prolactinomas1. Management recommendations include using the lowest dose for the shortest period possible to achieve these objectives, while monitoring for the development of flow murmurs at each visit. Echocardiographic evaluation should be considered in patients who require long-term treatment with cabergoline, especially in high doses. Guidelines for repeat echocardiographic intervals remain unclear. There is a need for larger, preferably prospective, studies with careful echocardiographic assessment and with longer durations of follow-up than the currently available studies.\n\nShould valvular regurgitation develop, early discontinuation and management with bromocriptine may be effective in reversing cardiac valvular dysfunction, as observed in this case study.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient. The OHSU institutional review board does not require additional consent for case reports.", "appendix": "Author contributions\n\n\n\nChris Yedinak, Troy H. Dillard, Kevin S. Wei, and Maria Fleseriu wrote the manuscript. Maria Fleseriu, Chris Yedinak, and Shirley McCartney revised the manuscript. Chris Yedinak and Kevin Wei selected the images. All authors approved the final manuscript for publication.\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\nMelmed S, Casanueva FF, Hoffman AR, et al.: Diagnosis and treatment of hyperprolactinemia: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2011; 96(2): 273–288. PubMed Abstract \| Publisher Full Text\n\nPeralta C, Wolf E, Alber H, et al.: Valvular heart disease in Parkinson's disease vs. controls: An echocardiographic study. Mov Disord. 2006; 21(8): 1109–1113. PubMed Abstract \| Publisher Full Text\n\nPinero A, Marcos-Alberca P, Fortes J: Cabergoline-related severe restrictive mitral regurgitation. N Engl J Med. 2005; 353(18): 1976–1977. PubMed Abstract \| Publisher Full Text\n\nColao A, Galderisi M, Di Sarno A, et al.: Increased prevalence of tricuspid regurgitation in patients with prolactinomas chronically treated with cabergoline. J Clin Endocrinol Metab. 2008; 93(10): 3777–3784. PubMed Abstract \| Publisher Full Text\n\nElenkova A, Shabani R, Kalinov K, et al.: Increased prevalence of subclinical cardiac valve fibrosis in patients with prolactinomas on long-term bromocriptine and cabergoline treatment. Eur J Endocrinol. 2012; 167(1): 17–25. PubMed Abstract \| Publisher Full Text\n\nTan LC, Ng KK, Au WL, et al.: Bromocriptine use and the risk of valvular heart disease. Mov Disord. 2009; 24(3): 344–349. PubMed Abstract \| Publisher Full Text\n\nAuriemma RS, Pivonello R, Perone Y, et al.: Safety of long-term treatment with cabergoline on cardiac valve disease in patients with prolactinomas. Eur J Endocrinol. 2013; 169(3): 359–366. PubMed Abstract \| Publisher Full Text\n\nBogazzi F, Buralli S, Manetti L, et al.: Treatment with low doses of cabergoline is not associated with increased prevalence of cardiac valve regurgitation in patients with hyperprolactinaemia. Int J Clin Pract. 2008; 62(12): 1864–1869. PubMed Abstract \| Publisher Full Text\n\nDrake WM, Stiles CE, Howlett TA, et al.: A cross-sectional study of the prevalence of cardiac valvular abnormalities in hyperprolactinemic patients treated with ergot-derived dopamine agonists. J Clin Endocrinol Metab. 2014; 99(1): 90–6. PubMed Abstract \| Publisher Full Text\n\nLancellotti P, Livadariu E, Markov M, et al.: Cabergoline and the risk of valvular lesions in endocrine disease. Eur J Endocrinol. 2008; 159(1): 1–5. PubMed Abstract \| Publisher Full Text\n\nBhattacharyya S, Schapira AH, Mikhailidis DP, et al.: Drug-induced fibrotic valvular heart disease. Lancet. 2009; 374(9689): 577–585. PubMed Abstract \| Publisher Full Text\n\nMast ST, Jollis JG, Ryan T, et al.: The progression of fenfluramine-associated valvular heart disease assessed by echocardiography. Ann Intern Med. 2001; 134(4): 261–266. PubMed Abstract \| Publisher Full Text\n\nBoguszewski CL, dos Santos CM, Sakamoto KS, et al.: A comparison of cabergoline and bromocriptine on the risk of valvular heart disease in patients with prolactinomas. Pituitary. 2012; 15(1): 44–49. PubMed Abstract \| Publisher Full Text\n\nZanettini R, Antonini A, Gatto G, et al.: Valvular heart disease and the use of dopamine agonists for Parkinson's disease. N Engl J Med. 2007; 356(1): 39–46. PubMed Abstract \| Publisher Full Text\n\nSchade R, Andersohn F, Suissa S, et al.: Dopamine agonists and the risk of cardiac-valve regurgitation. N Engl J Med. 2007; 356(1): 29–38. PubMed Abstract \| Publisher Full Text\n\nRuzicka E, Linkova H, Penicka M, et al.: Low incidence of restrictive valvulopathy in patients with Parkinson's disease on moderate dose of pergolide. J Neurol. 2007; 254(11): 1575–1578. PubMed Abstract \| Publisher Full Text\n\nJahnichen S, Horowski R, Pertz HH: Agonism at 5–HT2B receptors is not a class effect of the ergolines. Eur J Pharmacol. 2005; 513(3): 225–228. PubMed Abstract \| Publisher Full Text\n\nSerratrice J, Disdier P, Habib G, et al.: Fibrotic valvular heart disease subsequent to bromocriptine treatment. Cardiol Rev. 2002; 10(6): 334–336. PubMed Abstract\n\nDroogmans S, Roosens B, Cosyns B, et al.: Dose dependency and reversibility of serotonin-induced valvular heart disease in rats. Cardiovasc Toxicol. 2009; 9(3): 134–141. PubMed Abstract \| Publisher Full Text\n\nNachtigall LB, Valassi E, Lo J, et al.: Gender effects on cardiac valvular function in hyperprolactinaemic patients receiving cabergoline: a retrospective study. Clin Endocrinol (Oxf). 2010; 72(1): 53–58. PubMed Abstract \| Publisher Full Text\n\nVan Camp G, Flamez A, Cosyns B, et al.: Treatment of Parkinson's disease with pergolide and relation to restrictive valvular heart disease. Lancet. 2004; 363(9416): 1179–1183. PubMed Abstract \| Publisher Full Text\n\nValassi E, Klibanski A, Biller BM: Clinical Review#: Potential cardiac valve effects of dopamine agonists in hyperprolactinemia. J Clin Endocrinol Metab. 2010; 95(3): 1025–1033. PubMed Abstract \| Publisher Full Text\n\nVallette S, Serri K, Rivera J, et al.: Long-term cabergoline therapy is not associated with valvular heart disease in patients with prolactinomas. Pituitary. 2009; 12(3): 153–157. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5594", "date": "04 Aug 2014", "name": "Atanaska Elenkova", "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 is well written, treats an actual problem (the risk of development of valvulopathy after long-term cabergoline treatment in patients with macroprolactinoma) and provides evidence about the reversibility of valvular changes after timely discontinuation of DA treatment.Title and abstract: The title is appropriate for the content of the article. The abstract is concise and accurately summarizes the essential information of the paper although it would be better if the authors define more precisely the anatomic specificity of valvulopathy – mild mitral regurgitation. Case report: The clinical case presentation is comprehensive and detailed but there are some minor points that should be clarified:Please clarify the prolactin levels at diagnosis. In the Presentation section (line 3) “At presentation, prolactin level was found to be grater than 1000 ng/ml on diluted testing” but in the section describing the laboratory evaluation at diagnosis (line 7) “Prolactin level was 55 ng/ml”. Was the difference due to so called “hook effect”?\n\nFigure 1: In the text the follow-up MR imaging is indicated to be “after 10 months of cabergoline treatment”. However, the figures 1C and 1D represent 2 years post-treatment MR images. Please clarify. Figure 2: Echocardiograms 2A and 2B are defined as baseline but actually they correspond to the follow-up echocardiographic assessment at the 4th year of cabergoline treatment. Did the patient undergo a baseline (prior to dopamine agonist treatment) echocardiographic evaluation? If he did not, it should be mentioned as study limitation in the Discussion section. The mitral valve thickness was mentioned to be normal. Did the echographic examination visualize increased echogenicity (hyperechogenicity) of the mitral cusps? How could you explain the decrease of LV ejection fraction (from 60-65% to 50-55%) after switching from cabergoline to bromocriptine treatment and respectively its increase to 62% after doubling the bromocriptine daily dose? Was LV function estimated always by the same method during the follow-up? Final paragraph: Authors conclude that early discontinuation and management with bromocriptine may be effective in reversing cardiac valvular dysfunction. Even though, regular echocardiographic follow up should be considered in patients who are expected to be on long-term high dose treatment with bromocriptine regarding its partial 5-HT2b agonist activity.", "responses": [] }, { "id": "5591", "date": "06 Aug 2014", "name": "Niki Karavitaki", "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\nVery interesting case.", "responses": [] }, { "id": "5593", "date": "11 Aug 2014", "name": "Corin Badiu", "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 prolactinoma in a young man. The case is instructive in terms of cabergoline impact upon associated valvulopathy.Few changes are required in order to improve the presentation / educational outcomes of the manuscript.The case is a partial resistant prolactinoma, since prolactin values remained above normal level and required a high CAB dosage after 10 months of treatment. The clinical examination should specify Tanner pubertal stage at initial presentation, since it is mentioned after 2 years of treatment. What was the effect of this treatment on bone age (previously mentioned as delayed)? No data are presented about fertility evaluation / treatment. For a man evaluated between 21-25 years, this might be of importance. Concerning heart evaluation, presented at 4 years after beginning of treatment, the authors should mention if they did heart ultrasound ever before? Some subclinical aspects can be revealed by detailed regular heart evaluation, especially on patients on high doses of Cabergoline.Obviously, alternative treatments should be commented on. First and foremost surgical approach, when the tumor became microadenoma, is a very reasonable way to treat without significant side-effects. The authors should comment on this option.", "responses": [] } ]	1	https://f1000research.com/articles/3-171
https://f1000research.com/articles/3-167/v1	22 Jul 14	{ "type": "Case Report", "title": "Case Report: Group B Streptococcus meningitis in an adolescent ", "authors": [ "Roselle Vittorino", "Joyce Hui-Yuen", "Adam J. Ratner", "Amy Starr", "Teresa McCann", "Roselle Vittorino", "Joyce Hui-Yuen", "Adam J. Ratner", "Amy Starr" ], "abstract": "Streptococcus agalactiae (group B Streptococcus, GBS) usually colonizes the gastrointestinal and lower genital tracts of asymptomatic hosts, yet the incidence of invasive disease is on the rise. We describe a case of an 18 year old woman, recently diagnosed with lupus, who reported a spontaneous abortion six weeks prior to her hospitalization. She presented with fever, altered mental status, and meningeal signs, paired with a positive blood culture for GBS. Magnetic resonance imaging of her brain demonstrated an extra-axial fluid collection, and she was diagnosed with meningitis. She received prolonged intravenous antibiotic therapy and aggressive treatment for lupus, leading to clinical recovery. This case illustrates the importance of recognizing GBS as a potential pathogen in all patients presenting with CNS infection.", "keywords": [ "Group B Streptococcus", "meningitis", "lupus", "teenager" ], "content": "Introduction\n\nIn this case report we describe an interesting and unique patient with group B Streptococcus (GBS) meningitis following a spontaneous abortion without instrumentation. Although our patient was recently diagnosed with systemic lupus erythematosus (SLE), this chronic disease has not been shown to increase the chances of developing invasive GBS infection1. Although GBS can cause severe infections outside of the neonatal period, it rarely manifests as meningitis, accounting for less than 2% of all cases of invasive GBS in the United States and Spain2,3. There have been only two reported cases of GBS meningitis following elective termination with cervical dilatation and uterine evacuation, both presenting to a hospital in Detroit, Michigan4,5 and none to date following spontaneous abortion.\n\n\nCase report\n\nIn October, 2013, an 18-year-old African American young woman diagnosed in August of that year with SLE presented to a community hospital complaining of headache, neck and back pain for five days, and vomiting with non-bloody diarrhea for one day. Additionally, she reported right-sided weakness. Six weeks prior to this presentation, she met the criteria for a diagnosis of SLE with arthritis, a photosensitive malar rash, and serum serology positivity for anti-nuclear antibodies (1:5120) as well as anti-double stranded DNA and anti-ribonuclear protein antibodies. A regimen of oral prednisone (10 mg twice daily) and hydroxychloroquine (200 mg twice daily) was begun. In the weeks subsequent to the diagnosis and treatment of SLE, and prior to her hospitalization, the patient had a positive home pregnancy test, which was followed by an unusually heavy menstrual cycle and presumed spontaneous abortion. The patient did not seek medical attention, and therefore no uterine manipulation or evacuation was performed. The patient’s family history is significant for a maternal aunt and paternal cousin with SLE, and a maternal great-uncle who died from Libman-Sacks endocarditis.\n\nAt the time of the acute presentation, she was febrile to 38.9°C, oriented to self only, and had positive Kernig and Brudzinski signs. A computed tomography scan of the head revealed an extra-axial fluid collection in the right frontotemporal region, with a mild local mass effect, but no midline shift and normal ventricles. Intravenous vancomycin and ceftriaxone were initiated and the patient was transferred to our institution.\n\nOn arrival to our pediatric intensive care unit, the patient’s temperature was 38.8°C, blood pressure was 145/76 mmHg, heart rate was 70 beats per minute, respiratory rate was 22 breaths per minute, and oxygen saturation was 100% on room air. She was drowsy and irritable when aroused, complaining of continued severe headache and neck pain. Neurological exam was significant for photophobia and meningismus. The remainder of the neurological exam and the ocular exam were normal. Analysis of the cerebrospinal fluid (CSF) revealed 1525 white blood cells/µL (91% neutrophils), 65 red blood cells/µL, 252 mg/dL protein, and 56 mg/dL glucose. An MRI of the brain with and without contrast revealed a moderately sized extra-axial effusion with secondary inflammatory response (Figure 1 and Figure 2) and punctate diffusion abnormalities within the dependent portion of the lateral ventricles, consistent with pus or debris. The CSF culture was negative for bacteria, likely secondary to receipt of antibiotics prior to sampling, but the blood culture prior to antibiotic receipt grew GBS. Antibiotic susceptibility testing was not performed. CSF studies for other pathogens, including cryptococcal antigen, detection of herpes simplex and varicella viruses by PCR, and specific cultures for acid-fast bacilli and fungi, were negative. Upon identification of GBS, vancomycin and ceftriaxone were discontinued, and treatment was continued with intravenous penicillin G (4,000,000 units every 4 hours) for two weeks. The patient showed rapid improvement on antibiotic therapy and had resolution of all neurologic symptoms within one week of admission. She was discharged to receive intravenous ceftriaxone for one additional week at home.\n\nNotably, the patient’s SLE was severely active during the course of this acute illness. Initial laboratory results at our institution demonstrated anti-Smith, anti-Ro and anti-La antibodies. She also had severe hypocomplementemia, doubling of serum creatinine, and nephrotic range proteinuria as assessed by 24-hour urine collection. Subsequent kidney biopsy demonstrated class V lupus nephritis. Induction therapy with intravenous methylprednisolone was started, which was then changed to oral prednisone (30 mg twice daily), mycophenolate mofetil (1000 mg twice daily), and hydroxychloroquine (400 mg once daily). Her serologically active SLE began to improve after her course of methylprednisolone, but was still active at the time of discharge.\n\n\nDiscussion\n\nGBS is an encapsulated bacterium that produces a narrow zone of hemolysis on sheep blood agar. Strains are classified into one of 10 serotypes determined by the organism’s polysaccharide capsule. GBS colonizes the gastrointestinal tract, vagina, and urethra of asymptomatic hosts, but can cause invasive disease in infants, pregnant or post-partum women, individuals with underlying medical conditions and the elderly6–9.\n\nGBS emerged as the leading cause of neonatal sepsis in the 1970s in the US10. While the rates of early-onset invasive GBS infection among neonates have declined due to widespread use of intra-partum antibiotic prophylaxis in the United States, GBS continues to affect an estimated 1200 infants per year9. Pregnancy and the post-partum state are characterized by an altered immune response that allows GBS to cause invasive disease11. A 6-year epidemiologic study done in the US found that half of all invasive GBS infections seen in pregnant women were associated with infection of the upper genital tract, placenta, and amniotic sac, leading to fetal demise12. Other commonly reported manifestations included bacteremia (31%) and endometritis without fetal death (8%). Deutscher et al. reported that a greater proportion of cases of invasive GBS disease occurred in the post-partum period (14% vs. 10% in pregnant women), and pregnant women had a 2-fold increase incidence of GBS disease compared to non-pregnant women11.\n\nAs the rate of neonatal GBS infection has declined, the majority of cases now occur in the adult population, manifesting as skin or soft tissue infection, bacteremia, pneumonia, osteomyelitis and septic arthritis2,3,6,7,12. Several studies have shown increasing rates of invasive GBS infection in all adult age groups, but particularly vulnerable populations are those with medical co-morbidities and the elderly. Of those with underlying medical conditions, 41% have diabetes mellitus, 36% cardiovascular disease, and 17% malignancy12. Interestingly, immunosuppressed individuals comprise less than 10% of all at risk groups2,3.\n\nGBS meningitis is a rare occurrence, accounting for <2% of all cases of invasive GBS infection in the US and Spain2,3. In fact, one of the largest initial reviews of GBS meningitis in the adult population identified only 12 cases over a 15 year period, and found 72 cases in the literature overall. The majority of those cases were associated with underlying medical conditions and the risk of contracting GBS meningitis was greater with increasing age, with 58% of cases found in individuals >50 years old13. Although pregnant women have a 2-fold increased incidence of GBS disease compared to non-pregnant women, meningitis caused by GBS remains rare in this population14–18.\n\nInfection of the central nervous system (CNS) is uncommon among patients with SLE, and relevant symptoms may resemble neuropsychiatric lupus flare, making diagnosis difficult1. There does not appear to be a temporal association between contraction of CNS infection and diagnosis of SLE19, but CNS infection does account for a significant amount of mortality in patients with SLE20. To date, the largest cohort of SLE patients with CNS infections is a group of 38 patients, none of whom had GBS, collected over a 10-year period19. In a large retrospective review of Korean patients with SLE, only 1.4% had meningitis, with the leading causative organism being Cryptococcus neoformans21. A recent study characterizing the major infections seen in juvenile onset SLE found meningitis to be a rare entity. Only one case was identified out of 101 major infections, with the causative agent being Streptococcus pneumoniae22. Only three other cases of GBS meningitis in patients with SLE have been reported3,23.\n\nTo our knowledge, this patient represents the only reported case of GBS meningitis following a spontaneous abortion without instrumentation of the uterus. The two previously reported cases of GBS meningitis followed elective abortion with cervical dilatation and uterine evacuation4,5. Although these women became symptomatic within two weeks of their procedure, a recent study of expectant versus surgical management of first-trimester miscarriage showed that 18.6% of women in the expectant management group had evidence of retained products of conception after four weeks24. Therefore it is possible that our patient remained at risk for invasive infection longer than would otherwise be expected.\n\nThis case illustrates the importance of recognizing GBS as a potential pathogen in all patients presenting with CNS infection. It also highlights the value of obtaining a thorough obstetrical history in all reproductive-aged females, even in the setting of acute illness.\n\n\nConsent\n\nWritten informed consent for publication of the patient’s clinical details and images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nRV, JH, and TM equally contributed to the writing and editing of this manuscript. Additional editing and expert content was provided by AS and AR.\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\nYang CD, Wang XD, Ye S, et al.: Clinical features, prognostic and risk factors of central nervous system infections in patients with systemic lupus erythematosus. Clin Rheumatol. 2007; 26(6): 895–901. PubMed Abstract \| Publisher Full Text\n\nBlancas D, Santin M, Olmo M, et al.: Group B streptococcal disease in nonpregnant adults: incidence, clinical characteristics, and outcome. Eur J Clin Microbiol Infect Dis. 2004; 23(3): 168–173. PubMed Abstract \| Publisher Full Text\n\nSkoff TH, Farley MM, Petit S, et al.: Increasing burden of invasive group B streptococcal disease in nonpregnant adults, 1990–2007. Clin Infect Dis. 2009; 49(1): 85–92. PubMed Abstract \| Publisher Full Text\n\nDeziel PJ, McGuire N, Brown PD: Group B streptococcal meningitis complicating elective abortion: report of 2 cases. Clin Infect Dis. 2000; 31(5): E23–E25. PubMed Abstract \| Publisher Full Text\n\nWalker MA, McNeeley SG: Group B streptococcus meningitis following elective termination of pregnancy: two case reports. Infect Dis Obstet Gynecol. 1995; 2(6): 275–278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFarley MM: Group B streptococcal disease in nonpregnant adults. Clin Infect Dis. 2001; 33(4): 556–561. PubMed Abstract \| Publisher Full Text\n\nLe Doare K, Heath PT: An overview of global GBS epidemiology. Vaccine. 2013; 31(Suppl 4): D7–D12. PubMed Abstract \| Publisher Full Text\n\nOhlsson A, Shah VS: Intrapartum antibiotics for known maternal group B streptococcal colonization. Cochrane Database Syst Rev. 2013; 1: CD007467; 1–38. PubMed Abstract \| Publisher Full Text\n\nVerani JR, McGee L, Schrag SJ: Prevention of perinatal group B streptococcal disease--revised guidelines from CDC, 2010. MMWR Recomm Rep. 2010; 59(RR-10): 1–36. PubMed Abstract\n\nFranciosi RA, Knostman JD, Zimmerman RA: Group B streptococcal neonatal and infant infections. J Pediatr. 1973; 82(4): 707–18. PubMed Abstract \| Publisher Full Text\n\nDeutscher M, Lewis M, Zell ER, et al.: Incidence and severity of invasive Streptococcus pneumoniae, group A Streptococcus, and group B Streptococcus infections among pregnant and postpartum women. Clin Infect Dis. 2011; 53(2): 114–123. PubMed Abstract \| Publisher Full Text\n\nPhares CR, Lynfield R, Farley MM, et al.: Epidemiology of invasive group B streptococcal disease in the United States, 1999–2005. JAMA. 2008; 299(17): 2056–2065. PubMed Abstract \| Publisher Full Text\n\nDomingo P, Barquet N, Alvarez M, et al.: Group B streptococcal meningitis in adults: report of twelve cases and review. Clin Infect Dis. 1997; 25(5): 1180–1187. PubMed Abstract \| Publisher Full Text\n\nGhani NA, Jaafar R, Ishak S, et al.: Mother with post-partum group B Streptococcus meningitis and cerebellar abscess. J Obstet Gynaecol Res. 2007; 33(2): 195–198. PubMed Abstract \| Publisher Full Text\n\nGrossman J, Tompkins RL: Group B beta-hemolytic streptococcal meningitis in mother and infant. N Engl J Med. 1974; 290(7): 387–388. PubMed Abstract \| Publisher Full Text\n\nGuerin JM, Leibinger F, Mofredi A, et al.: Streptococcus B meningitis in post-partum. J Infect. 1997; 34(2): 151–153. PubMed Abstract \| Publisher Full Text\n\nRavindranth NT, O’Driscoll J: Maternal group B streptococcal meningitis in the postpartum period. J Obstet Gynaecol. 2003; 23(1): 79. PubMed Abstract\n\nWolfe RR Jr, Norwick ML, Bofill JA: Fatal maternal beta-hemolytic group B streptococcal meningitis: a case report. Am J Perinatol. 1998; 15(11): 597–600. PubMed Abstract \| Publisher Full Text\n\nBenseler SM, Silverman ED: Neuropsychiatric involvement in pediatric systemic lupus erythematosus. Lupus. 2007; 16(8): 564–71. PubMed Abstract \| Publisher Full Text\n\nZonana-Nacach A, Yañez P, Jiménez-Balderas FJ, et al.: Disease activity, damage and survival in Mexican patients with acute severe systemic lupus erythematosus. Lupus. 2007; 16(12): 997–1000. PubMed Abstract \| Publisher Full Text\n\nKim JM, Kim KJ, Yoon HS, et al.: Meningitis in Korean patients with systemic lupus erythematosus: analysis of demographics, clinical features and outcomes; experience from affiliated hospitals of the Catholic University of Korea. Lupus. 2011; 20(5): 531–536. PubMed Abstract \| Publisher Full Text\n\nCosta-Reis P, Nativ S, Isgro J, et al.: Major infections in a cohort of 120 patients with juvenile-onset systemic lupus erythematosus. Clin Immunol. 2013; 149(3): 442–449. PubMed Abstract \| Publisher Full Text\n\nKim BS, Kim SH, Kim SJ, et al.: An unusual complication of systemic lupus erythematosus: bacterial meningitis caused by Streptococcus agalactiae. J Eur Acad Dermatol Venereol. 2007; 21(9): 1274–1275. PubMed Abstract \| Publisher Full Text\n\nAl-Ma’ani W, Solomayer EF, Hammadeh M: Expectant versus surgical management of first-trimester miscarriage: a randomised controlled study. Arch Gynecol Obstet. 2014; 289(5): 1011–1015. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5546", "date": "04 Aug 2014", "name": "Theresa T. Lu", "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 case report about a newly diagnosed lupus patient who develops GBS meningitis after a spontaneous abortion. Most of the salient details are included and there is an informative review of the literature on GBS infections, autoimmune disease, and spontaneous abortions. This is a useful documentation of an unusual case that can be helpful when considering the differential diagnosis of meningitis in a patient with lupus, other autoimmune disorders, or spontaneous abortion.One presumes that the patient was in her first trimester, but confirmation of that in the description of the patient would be helpful to be complete.", "responses": [] }, { "id": "5871", "date": "10 Oct 2014", "name": "Rebecca Madan", "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 case report reviews an unusual and thought-provoking case of Group B Streptococcus meningitis in a young woman who had a recent history of both spontaneous abortion and newly diagnosed SLE. Of note, she received relatively low-dose immunosuppression prior to developing meningitis, and the authors indicate that SLE participants are typically not thought to be at higher risk for bacterial meningitis. This case report is highly relevant in that it illustrates the imperative to 1) exclude infection in the setting of altered mental status in SLE patients and 2) to investigate for potential sources of infection (recent pregnancy) in a patient with an invasive bacterial infection that is not typical of their age group.The authors imply in their Discussion that this case of GBS meningitis was more likely to be related to the patient's recent history of spontaneous abortion and was less likely related to her recently diagnosed SLE. They cite a significant prevalence of endometritis (8%) in women who develop invasive GBS with pregnancy. Did the patient have evidence of endometritis on exam or imaging? Did she undergo pelvic imaging to assess for retained products of conception (or undisclosed foreign body) that could provide an ongoing source of infection?The most interesting aspect of this case is this patient's significant SLE flare in the setting of GBS meningitis. The authors report that she had serologic evidence of SLE flare, hypocomplementemia, and a new diagnosis of class V lupus nephritis concomitant with her GBS diagnosis. This presents an interesting finding to expand upon in the Discussion. Although bacterial meningitis is unusual in the setting of SLE, is it possible that this young woman developed a severe flare prior to meningitis, and that the subsequent immune dysfunction predisposed her to invasive GBS? Or do the authors postulate that immune dysfunction in the setting of bacterial sepsis resulted in the SLE flare? What are the possible mechanisms that could underly a relationship between these two conditions? A brief discussion of the patient's SLE flare would contribute significantly to this manuscript.", "responses": [] } ]	1	https://f1000research.com/articles/3-167
https://f1000research.com/articles/2-221/v1	21 Oct 13	{ "type": "Research Article", "title": "Resolving the paradox for protein aggregation diseases: a common mechanism for aggregated proteins to initially attack membranes without needing aggregates", "authors": [ "Haina Qin", "Liangzhong Lim", "Yuanyuan Wei", "Garvita Gupta", "Jianxing Song", "Haina Qin", "Liangzhong Lim", "Yuanyuan Wei", "Garvita Gupta" ], "abstract": "Paradoxically, aggregation of specific proteins is characteristic of many human diseases and aging, yet aggregates have been found to be unnecessary for initiating pathogenesis. Here we determined the NMR topology and dynamics of a helical mutant in a membrane environment transformed from the 125-residue cytosolic all-β MSP by the ALS-causing P56S mutation. Unexpectedly, despite its low hydrophobicity, the P56S major sperm protein (MSP) domain becomes largely embedded in the membrane environment with high backbone rigidity. Furthermore it is composed of five helices with amphiphilicity comparable to those of the partly-soluble membrane toxin mellitin and α-synuclein causing Parkinson's disease. Consequently, the mechanism underlying this chameleon transformation becomes clear: by disrupting the specific tertiary interaction network stabilizing the native all-β MSP fold to release previously-locked amphiphilic segments, the P56S mutation acts to convert the classic MSP fold into a membrane-active protein that is fundamentally indistinguishable from mellitin and α-synuclein which are disordered in aqueous solution but spontaneously partition into membrane interfaces driven by hydrogen-bond energetics gained from forming α-helix in the membrane environments. As segments with high amphiphilicity exist in all proteins, our study successfully resolves the paradox by deciphering that the proteins with a higher tendency to aggregate have a stronger potential to partition into membranes through the same mechanism as α-synuclein to initially attack membranes to trigger pathogenesis without needing aggregates. This might represent the common first step for various kinds of aggregated proteins to trigger familiar, sporadic and aging diseases. Therefore the homeostasis of aggregated proteins in vivo is the central factor responsible for a variety of human diseases including aging. The number and degree of the membrane attacks by aggregated proteins may act as an endogenous clock to count down the aging process. Consequently, a key approach to fight against them is to develop strategies and agents to maintain or even enhance the functions of the degradation machineries.", "keywords": [ "Protein aggregation/insolubility is characteristic of a broad spectrum of human diseases", "in particular neurodegenerative/aging diseases1", "2", "which include Parkinson’s disease (PD)", "Alzheimer’s disease (AD)", "Huntington’s disease (HD)", "spinocerebellar ataxias (SCA)", "and amyotrophic lateral sclerosis (ALS). In addition", "protein aggregation has been shown to play a role in aging3 as well as cardiomyocyte autophagy4 and type II diabetes5–7. Remarkably", "for the above mentioned diseases", "aggregation/insolubility of specific proteins can be triggered by either genetic mutations (familiar) or environmental insults (sporadic)", "which strongly implies that a common mechanism may exist to initiate both familiar and sporadic forms of these clinically distinct diseases. Paradoxically", "recent studies have suggested that the accumulation of aggregates is unlikely to be the first step in pathogenesis7–9. However", "the common mechanism to initiate these diseases still remains to be elucidated1", "7–9." ], "content": "Introduction\n\nProtein aggregation/insolubility is characteristic of a broad spectrum of human diseases, in particular neurodegenerative/aging diseases1,2, which include Parkinson’s disease (PD), Alzheimer’s disease (AD), Huntington’s disease (HD), spinocerebellar ataxias (SCA), and amyotrophic lateral sclerosis (ALS). In addition, protein aggregation has been shown to play a role in aging3 as well as cardiomyocyte autophagy4 and type II diabetes5–7. Remarkably, for the above mentioned diseases, aggregation/insolubility of specific proteins can be triggered by either genetic mutations (familiar) or environmental insults (sporadic), which strongly implies that a common mechanism may exist to initiate both familiar and sporadic forms of these clinically distinct diseases. Paradoxically, recent studies have suggested that the accumulation of aggregates is unlikely to be the first step in pathogenesis7–9. However, the common mechanism to initiate these diseases still remains to be elucidated1,7–9.\n\nALS is the most prevalent fatal motor neuron disease, yet its underlying mechanism still remains a mystery despite intense studies since the first description more than 130 years ago10. Approximately 10% of ALS cases have a hereditary background, while the other cases are sporadic10. ALS8 was identified from a large Brazilian family, and encodes a mutated P56S major sperm protein MSP domain of VAPB (vesicle-associated membrane protein-associated protein B)11. In the cytosol, the 125-residue MSP domain adopts a seven-stranded immunoglobulin-like β-sandwich fold (Figure 1A), which is anchored onto the endoplasmic reticulum (ER) surface (Figure 1B)12. The MSP domain can also be cleaved from its transmembrane anchor to serve as a ligand for the EphA4 receptor1,14, which is the only-known ALS modifier15. Noticeably, inhibition of EphA4 by a small molecule, called C1, which targets the EphA4 ligand binding channel16,17 rescued the disease phenotype in ALS models15.\n\nA. 125-residue wild-type MSP domain adopting a seven-stranded immunoglobulin-like β-sandwich fold, with Pro56 displayed in spheres. B. The wild-type MSP domain of VAPB is anchored onto the ER membrane facing the cytosol by a C-terminal transmembrane fragment. C. The ALS-causing P56S mutant is able to remodel ER to have stacked cisternae by acquiring ability of the P56S MSP to interact with membranes. D. Far-UV CD spectrum of the wild-type MSP domain (black), typical of a β structure; and spectra of the P56S MSP in aqueous solution (cyan); in DMPC vesicle (green), bicelle formed by DMPC and DHPC (blue) as well as in DPC micelle (red) at pH 4.0. E. Far-UV CD spectra of the P56S MSP in DMPC vesicle (purple), bicelle formed by DMPC and DHPC (green) and in DPC micelle (blue) in 5 mM phosphate buffer at pH 7.5.\n\nThe ALS-causing P56S mutation renders VAPB to form detergent-resistant aggregates in vivo upon overexpression18. In vitro, we have shown that indeed the P56S MSP mutant is completely insoluble in buffers12. Nevertheless, our unique discovery that all insoluble proteins, including the most hydrophobic integral membrane peptide, could be dissolved in unsalted water and manifest their intrinsic conformations19–22 allowed us to characterize the residue-specific conformation of the P56S MSP domain in aqueous solution by NMR spectroscopy12. Remarkably, we showed that the P56S mutation is sufficient to completely eliminate its native β-sandwich fold and consequently the P56S MSP domain becomes predominantly-disordered, only with weakly-populated helical conformations over several regions. As such, both in vivo and in vitro results highlight the association of the aggregation of the P56S mutant with the ALS pathogenesis.\n\nOn the other hand, a recent study failed to detect any significant accumulation of aggregates in motor neurons derived from induced pluripotent stem cells of patients carrying the P56S mutation23, suggesting that the accumulation of the P56S VAPB aggregates is not the primary trigger for ALS8 pathogenesis. Furthermore, two recent studies showed that the P56S mutant acquired a novel ability to remodel the endoplasmic reticulum (ER) to have stacked cisternae even without needing the accumulation of aggregates/inclusions24,25. On the other hand, we discovered that the unstructured P56S, but not wild-type MSP domain, is able to insert into a membrane environment to become a helical structure26, thus providing the underlying mechanism (Figure 1C) for the observation22,24,25.\n\nTo shed light on how a point mutation can transform a well-folded, all-β domain into a helical membrane protein, as well as understanding the role of this transformation in initiating ALS pathogenesis, here by solution NMR spectroscopy and paramagnetic relaxation enhancement (PRE), we determined the three-dimensional topology and dynamics of the 125-residue P56S MSP domain in a membrane environment. This represents the first three-dimensional topology of the membrane-embedded helical proteins which are transformed from a well-folded cytosolic all-β domain. Astonishingly, the P56S MSP domain is mostly embedded in the membrane environment with high backbone rigidity, and is composed of five well-formed helices at N- and C-ends linked by a long unstructured loop. Although no membrane-associated fragments could be detected based on hydrophobicity used for identifying classic membrane proteins, the helical residues were found to possess high amphiphilicity that was comparable to those of the membrane-active toxin mellitin and the intrinsically-unstructured α-synuclein that cause Lewy body diseases. This immediately reveals the mechanism for the chameleon transformation: the P56S mutation acts to convert the well-folded cytosolic MSP domain into an unstructured membrane-active protein like mellitin and α-synuclein, by disrupting the specific long-range interaction network that stabilizes the native β-sandwich MSP fold12. Consequently, the previously locked intrinsic amphiphilic and other hydrophobic regions are released and accessible to bulk solvent, which leads to severe aggregation in buffers but, on the other hand, also drives partition into membranes. Since we, and others have extensively shown that insoluble proteins lack tight tertiary packing19–22,27,28; and segments with high intrinsic amphiphilicity universally exist in all proteins including random sequences, regardless of their native structures29,30, our current study thus resolves the paradox by deciphering that all disease-associated proteins, regardless of being partly-soluble like α-synuclein or insoluble like the P56S MSP, share a common mechanism to attack membranes without needing aggregates. This mechanism might represent the initial step in triggering familiar, sporadic and aging diseases.\n\n\nResults\n\nWe first accessed the conformational properties in different environments by circular dichroism (CD) spectroscopy. As shown in Figure 1D, the wild-type MSP domain has a far-UV CD spectrum typical of a β-sheet protein. The P56S mutant is predominantly disordered, without any stable secondary structure in aqueous solution, as we previously reported12. Strikingly, the P56S MSP domain transforms into similar helical conformations in 1,2- 1,2-DMPC (dimyristoyl-sn-glycero-3-phosphocholine) vesicles, bicelles formed by DMPC/DHPC (1,2-dihexanoyl-sn-glycero-3-phosphocholine), as well as DPC (n-dodecylphosphocholine) micelles (Figure 1D). Although the P56S MSP domain gets aggregated immediately in buffers12, once inserted into membranes, it adopts similar helical conformations even in the presence of phosphate buffer at pH 7.5 (Figure 1E).\n\nBy extensively screening lipid component, solution and temperature conditions for NMR experiments, we succeeded in acquiring and subsequently analysing a large set of high-quality NMR spectra in DPC micelles. While consistent with CD data, the NMR chemical shift index31,32 demonstrates that in aqueous solutions the P56S MSP domain is highly unstructured, and only has weakly-populated helical conformations over several regions. Upon partitioning into the membrane environment, five regions have very large (ΔCα-ΔCβ) chemical shifts comparable to those expected for the well-formed helix, unambiguously showing the formations of stable helices over Lys3-Val7, Phe22-Leu30, Val90-Met93, Asp98-Lys107 and Asp116-Leu125 (Figure 2A). On the other hand, there is no region retaining the native β-sheet secondary structure. The formation of the helices is further supported by the extensive manifestation of NOEs defining the helical structure, which include dNN(i, i+1), dαN(i, i+2), dαN(i, i+3), and dαN(i, i+4) (Figure 2D). Amazingly, there is a long region over residues Gly33-His86 without significant changes of chemical shifts upon partitioning into the DPC micelle, indicating that even in the membrane environment this region remains largely unstructured as in aqueous solution.\n\nA. Residue specific (ΔCα-ΔCβ) values of the P56S MSP in aqueous solution (blue) and in DPC micelle (red). The blue arrows are used for indicating the β-strands in the wild-type MSP structures and red cylinders for helices formed in DPC micelles. B. NOE connectivities defining secondary structures of the P56S MSP in DPC micelle. The seven residues selected for spin-labeling are colored in red.\n\nBy analyzing 15N- and 13C-edited NOESY spectra of the P56S MSP domain in both DPC and deuterated DPC, we identified a large set of NOEs defining the α-helices but only very limited long-range NOEs. Thus, to define its three-dimensional topology, we introduced the free radical probe, MTSSL at seven sites as indicated in Figure 2B. Subsequently we utilize paramagnetic relaxation enhancement (PRE) to obtain long distance constraints by the well-established approach33–35. Finally using X-PLOR and CNS36,37, we calculated the three-dimensional topology of the P56S MSP domain in a DPC micelle with experimental constraints including distances derived from 339 sequential, 162 medium-range and 7 long-range NOEs; and 465 PREs, as well as 59 pairs of phi and psi dihedral angles predicted by TALOS (http://spin.niddk.nih.gov/bax/nmrserver/talos/)38.\n\nFigure 3A presents the superimposition of the 10 lowest-energy structures which are composed of five well-formed helices over residues 3–7, 22–30, 90–94, 98–107 and 116–125, consistent with NMR chemical shifts and NOE patterns (Figure 2). The long region over residues Gly33-His86 has no well-formed secondary structure, only with helices over Ile61-Val71 in two structures. The P56S mutation is located in the unstructured loop (Figure 3B). In all 10 structures, the orientation among the five helices is well-defined, with average RMS deviations of 1.9 Å for all atoms; 1.6 Å for heavy atoms and 0.9 Å for backbone atoms if only superimposed over the five helices. This indicates that the incorporation of PRE-derived long-range distances into the structure calculation is indeed a very effective approach to define the overall topology, as extensively demonstrated33–35. Noticeably, backbone hydrogen bonds are extensively formed within the helices (Figure 3C, 3D and 3E). This observation supports the previous notion that as in membrane environments, proteins are significantly shielded from the water molecules which have strong capacity to form intermolecular hydrogen bonds with protein atoms, proteins thus acquire strong ability to form intra-hydrogen bonds, thus favouring the formation of helix secondary structures. This so called “hydrogen-bond energetic” in fact represents a main force to drive the partition of amphiphilic proteins like mellitin into membranes39–42.\n\nA. Superimposition of the 10 selected NMR structures of the P56S MSP in DPC micelles. B. The lowest-energy structure. C. Two helices formed at the N-terminus. D. Three helices formed at the C-terminus; and E. Helices formed in the middle only found in two NMR structures. The purple dashed lines are used to indicate hydrogen bonds. F. Ribbon and G. surface representations of the P56S MSP in DPC micelle with the display of residues accessible to Mn2+ (green), to gadodiamide (blue). Yellow is used to color Pro residues and purple for residues with missing HSQC peaks.\n\nUnlike classic membrane proteins, no tight tertiary packing exists in the membrane-embedded P56S MSP domain, most likely due to the fact that it is transformed from a cytosolic all-β protein and therefore owns no specific tertiary interactions acquired in evolution for the classic membrane proteins. As a consequence, it represents a nice example of a protein in which folding can indeed stop in the middle of the stepwise folding models, namely at the formation of secondary structures40,43, thus highlighting the indispensable role of specific long-range interactions in specifying the tertiary structure of membrane proteins. The loose tertiary packing can in fact offer an advantage to rearrange the tertiary topology but to retain very similar secondary structures (Figure 1E) in different membranes as different lipids have been shown to poses no significant effects on the formation of the helix44. In fact, some of such non-classic properties such as presence of unstructured loops within membranes are starting to be observed even in classic membrane proteins45.\n\nWe also used HSQC titrations with two paramagnetic agents, gadodiamide and Mn2+, to probe the exposure of the P56S MSP domain in the DPC micelle35. Interestingly, 17 backbone amide protons are accessible to gadodiamide (Figure 3F), indicating that only a small portion of residues are exposed to bulk water and therefore the P56S MSP is mostly embedded in the membrane environment. Furthermore, 30 extra backbone amide protons are accessible to Mn2+, suggesting that these residues are located in the polar head-group phase of DPC micelle. As such, ~60% residues are possibly buried in the non-polar hydrocarbon phase, or/and involved in forming hydrogen bonds, which include the N-terminal second helix over residues Phe22-Leu30 and a large portion of unstructured loop (Figure 3F and 3G).\n\nTo pinpoint the backbone dynamic properties of the P56S MSP in aqueous solution and in the membrane environment, we acquired the heteronuclear NOE which reflects the backbone motions on the ps-ns time scale20,46–48. In aqueous solution, very small hNOEs were observed on the P56S MSP residues with an average of 0.08, and several N-terminal residues even had negative hNOE (Figure 4A and 4B), indicating that the P56S MSP residues are very flexible in aqueous solution. In contrast, once embedded in the membrane environment, all P56S MSP residues have positive hNOE values, with an average of 0.71. In particular, the C-terminal residues forming the helix had hNOEs reaching 1, which was comparable to those observed on any well-folded proteins46–48. Strikingly, although in general residues forming helices have larger hNOEs, the unstructured regions over Gly33-His86 also have hNOEs much larger than the corresponding residues in aqueous solution, with an average of 0.6. This observation is consistent with the titration results by gadodiamide and Mn2+ that the majority of the P56S MSP residues are embedded in membrane environment, which thus have highly restricted backbone motions on the ps-ns time scale, even without any regular secondary structure.\n\nA. {1H}–15N heteronuclear steady-state NOE (hNOE) of the P56S MSP in aqueous solution (blue) and in a DPC micelle (red). B. Structure of the P56S MSP in DPC micelles, with Pro residues colored in yellow; missing or overlapped residues in green and residues with hNOEs > the average in red. C. Difference of effective transverse relaxation rate R2(τcp) at 80 and 960 Hz. Inlet: dispersion curves for two residues S70 and F76. Red cylinders are used to indicate helices formed in DPC micelles. D. Structure of the P56S MSP in a DPC micelle, with Pro residues colored in yellow; residues missing or with data having large noise in green; and residues with ΔR2(τcp) > 2 in pink and residues with ΔR2(τcp) > 6 in red.\n\nWe further used 15N CPMG relaxation dispersion experiments to assess the backbone motions on the µs-ms time scale49. While the P56S MSP in aqueous solution had no detectable backbone motions on the µs-ms time scale (data not shown), many P56S MSP residues had backbone motions on the µs-ms time scale in the membrane environment (Figure 4C and 4D). In particular, significant conformational exchanges could be observed over residues Val44-Thr47 and Ser66-Phe76. The disappearance of HSQC peaks for residues Val71-Gln74 is likely due to large conformational exchanges on the µs-ms time scale. Unfortunately we have collected the CPMG relaxation dispersion data at 500 MHz but the quality is very poor. As such, only based on the data at one field (800 MHz), we were not able to fit the data to obtain parameters for the conformational exchanges. As many important biological events occur on the µs-ms time scale, the existence of significant µs-ms motions in the P56S MSP domain might impose considerable perturbations/damage to the biological functions of the membranes.\n\n\nDiscussion\n\nAll living cells and organelles in eukaryotic cells are surrounded by biological membranes. Most biological membranes are not only composed of phospholipids, but contain a large fraction of proteins embedded within the lipids. This protein fraction is estimated to make up half of the mass of a biological membrane50. Membrane proteins play various key roles in essential biological processes including cell signalling, transport of membrane-impermeable molecules and intercellular communication. Consequently, membrane proteins constitute the largest class of drug targets51. These classic membrane proteins have a high hydrophobicity, which plays a predominant role in their membrane-insertion, folding and stabilization52,53. By calculating hydrophobicity, a genome-wide analysis revealed that 20–30% of the open reading frames (ORFs) of various genomes encode integral helix-bundle membrane proteins54.\n\nHere, we determined the three-dimensional topology of a membrane-embedded helical protein which is transformed from a cytosolic all-β domain, triggered by an ALS-causing P56S mutation (Figure 5A and 5B). Unexpectedly, five well-formed helices in the membrane environment adopted β-strands in the native MSP fold (Figure 5C–5J). Based on the hydrophobicity55,56, which accurately predicts the transmembrane helix at the VAPB C-terminus, no membrane-associated helix can be detected within either wild-type or P56S MSP domain (Figure 5K and 5L). Nonetheless, calculation of the hydrophobic moment41,57 revealed that the helical residues have high amphiphilic α-helix potential29,30, which include Phe22-Leu32 and Cys53-Ala63, Asn68-Val69, Ser92, Lys107 and Asp116-Leu125 (Figure 5M). In particular, two regions over Phe22-Leu32 and Asp116-Leu125 have amphiphilicity comparable to mellitin, a honeybee membrane-active toxin39,58–60, and the intrinsically-unstructured α-synuclein that triggers Parkinson's disease61–65. Both of them are unstructured in an aqueous solution, and have a high tendency to aggregate, but spontaneously insert into membranes to form amphiphilic α-helices. Therefore, the mechanism for the chameleon transformation becomes clear: by eliminating the well-folded all-β MSP fold to release previously locked amphiphilic and other hydrophobic patches (Figure 6A), the P56S mutation acts to convert the cytosolic MSP domain into a mellitin- and α-synuclein-like membrane-active protein (Figure 6B), which has an even higher tendency to aggregate in buffers (Figure 6C), but shares the potential to partition into membrane interfaces (Figure 6D) driven by hydrogen-bond energetics resulting from forming helix39–42. Therefore the insoluble P56S MSP is fundamentally indistinguishable from partly-soluble mellitin and α-synuclein, designated here as “dynamic membrane proteins”, but it is significantly different from classic membrane proteins in two aspects: 1) it has lower hydrophobicity and therefore amphiphilicity is expected to considerably contribute to its insertion, folding and stabilization in the membrane environments, and 2) it lacks the tight tertiary packing which may thus allow partitioning into different membranes by rearranging its tertiary topology, but retaining the similar secondary structures. The “dynamic membrane proteins” exemplified by mellitin and α-synuclein are extensively characteristic of these non-classic properties. Early studies of the bacterio-rhodopsin structure suggested that membrane proteins are “inside-out”. In other words, they consist of a hydrophilic interior and a hydrophobic exterior66. However, further studies indicate that this rule is not generally applicable even to classic membrane proteins40,66–68. Also there appears to have no major driving force to bury polar residues within the protein interior66. The energetic cost for inserting polar groups into a lipid environment is not that high if based on the biological hydrophobicity scale, rather than on other hydrophobicity scales, most of which were derived by utilizing apolar solvents40,66–68.\n\nA–B. Chameleon transformation from a seven-stranded immunoglobulin-like β-sandwich fold to a membrane-embedded helical structure. C–J. Secondary structures of different regions adopted in the wild-type and membrane-embedded P56S MSP respectively. K. Transmembrane scale of the full-length wild-type (blue) and P56S (red) VAPB (243 residues) calculated with the previous method56. L. Hydrophobicity scale calculated with the previous method55. M. Hydrophobic moment of the full-length wild-type (blue) and P56S (red) VAPB (243 residues) calculated with the previous method57. Hydrophobic moment of the honeybee membrane toxin mellitin is colored in black and that for α-synuclein in cyan.\n\nA. The wild-type protein domain like MSP adopts a well-folded three-dimensional structure and therefore its intrinsic amphiphilic and other hydrophobic patches are locked and shielded from being accessible to bulk solvent, thus being highly-soluble in salted aqueous solution. B. Some mutations on a well-folded protein like the ALS-causing P56S one are sufficient to completely eliminate its native fold. This results in improper exposure of the intrinsic amphiphilic and other hydrophobic patches. The protein then becomes only soluble but unstructured in unsalted aqueous solution, but is aggregated in vivo with ~150 mM ion concentrations (C). D. The unstructured mutant acquires the ability to spontaneously partitions into membranes driven by hydrogen bond energetics resulting from forming an amphiphilic helix. As we have shown that insoluble proteins are only insoluble in buffers but soluble in unsalted aqueous solution, even the aggregated mutant is able to partition into the membrane upon having access to membranes under some conditions. However, under normal physiological conditions, aggregates may be immediately detected and subsequently removed by different degradation machineries including ubiquitin-proteasome (E) or/and autophagosome-lysosome (F) pathways. However, once these machineries become dysfunctional due to aging, or/and are inhibited by abnormal conditions, which are generally found to trigger neurodegenerative diseases, aggregates may get become accumulated, which may increase the chance of them accessing/attacking membranes.\n\nSystematic studies disclosed a surprising fact. Segments with high amphiphilicity exist in all proteins, including randomly-generated sequences regardless of their native structures29,30. Indeed, nature has exploited a variety of polypeptides, including mellitin, with high amphiphilicity, to achieve antimicrobial, antifungal, antiviral, or anticancer activities by attacking biological membranes to modulate the structural and dynamical properties of the lipids on different length- and time scales69. Interestingly, partly-soluble proteins, causing various human diseases, have also been shown to attack membranes by transforming their unstructured states in aqueous solution, to amphiphilic helixes in membranes. These proteins include: prions of spongiform transmissible encephalopathies70,71, amyloid beta-(1–40) and beta-(1–42) peptides of Alzheimer's disease72,73, tau tangles of Alzheimer's disease74, α-synuclein of Parkinson's disease62,64,65, huntingtin of Huntington's disease75–77 and the islet amyloid polypeptide of type II diabetes5,7.\n\nFacilitated by our unique discovery that previously-thought insoluble proteins are only buffer-insoluble but in fact soluble in unsalted aqueous solution19–22, we have recently discovered that all of the insoluble proteins we tested were able to interact with membranes to different degrees22. Here, the determination of secondary structures and three-dimensional topology of the buffer-insoluble P56S MSP mutant in a membrane environment showed that it is fundamentally indistinguishable from partly-soluble α-synuclein and other disease-causing aggregated proteins. Furthermore, we also found that although the ALS-causing T46I mutation does not eliminate, but only destabilizes the MSP fold. However, the T46I mutant appears to also have amphiphilic and other hydrophobic patches more accessible to bulk solvent than the wild-type MSP, thus leading to aggregation in buffers as well as transformation into a helical conformation like the P56S MSP at high DPC concentrations14,22. Therefore, together with previous results, our present study establishes that aggregated proteins causing diseases, regardless of being partly-soluble or insoluble in buffers, share a common mechanism to initially attack membranes without the need to form aggregates. An interesting question thus arises if all proteins contain amphiphilic segments, why are aggregated proteins closely associated with various human diseases? Firstly, for well-folded proteins like the wild-type MSP domain, their surface residues are hydrophilic while hydrophobic/amphiphilic segments are locked inside, thus inaccessible to interacting with membranes. Secondly, for partially-folded or unfolded proteins, our results from characterizing insoluble proteins reveal that insolubility/aggregation in buffers is mostly due to the improper exposure of hydrophobic patches including those in amphiphilic regions. Therefore, the high tendency of a protein to aggregate reflects that it has highly-accessible hydrophobic or/and amphiphilic patches, which are also driving forces to partition it into membrane interfaces. In other words, the factors for driving aggregation in buffers and partitioning into membranes are at least partly overlapped. Consequently the paradox is resolved: proteins with a higher tendency to aggregate have stronger intrinsic potential to partition into membranes but the formation/accumulation of aggregates is not a prerequisite for this initial interaction with membranes. The accumulation of these proteins in membranes will lead to the formation of channel/aggregates/amyloid fibrils as previously proposed7,8,69,78–81.\n\nThe ability of aggregated proteins to strongly interact with membranes implies that their primary/first step to initiate pathogenesis might be to modulate the structural and dynamical properties of the lipids by a variety of mechanisms that have already been proposed7,8,69,78–81. As implied from our results here, the numbers of dynamic membrane proteins in cells are much larger than previously recognized. Cellular membranes may therefore be under constant attacking by these proteins, thus rationalizing the observation that most aggregation causing diseases are neurodegenerative diseases and aging as neurons such as cortical neurons are rarely replaced82. Once they get damaged, serious phenotypes will manifest. This may also explain why plants have no aging. Further formation/accumulation of aggregates may radically impose physical stresses/damages to membranes as well as on whole cells, which may be required for pathogenesis of some diseases. As a consequence, the tissue-specific expression of aggregated proteins may be one main factor in manifesting disease phenotypes. On the other hand, the wild-type proteins as exemplified by VAPB, whose mutants become aggregated and cause diseases, physiologically functions as enzymes or signalling components. As a result, the loss of these functions due to mutations may lead to disease-specific phenotypes as observed for the VAPB-MSP domain12–15. Probably, unlike the P56S VAPB mutant which can be delivered to the ER without requiring the formation of aggregates due to the presence of the C-terminal ER-anchoring helix24,25, most cytosolic insoluble mutants will not be able to access membranes under normal physiological conditions as they get aggregated immediately after synthesis and subsequently degraded by complex machineries (Figure 6E and 6F). Moreover, it also appears challenging to detect the initial interaction between these cytosolic aggregated proteins with membranes at the early stage because most of them may not cause significant morphological changes in cells. Only upon proteasomal inhibition, a condition commonly found in neurodegenerative diseases79, do aggregated proteins such as VAPB3, an insoluble splicing variant of VAPB without an ER-anchored region, accumulate and thus have opportunities to access and attack membranes, and thus lead to sporadic diseases. Indeed, an increased expression of the wild-type α-synuclein due to gene duplication and triplication is required to initiate Parkinson's disease61,63.\n\nStrikingly, as illustrated in Figure 7, eukaryotic, particularly human, genomes appear to contain many pre-existing proteins like VAPB3, which have no intrinsic ability to fold into well-defined structures and consequently will unavoidably aggregate in vivo22. On the other hand, some family members or individuals carry additional genetic mutants like the P56S-VAPB which are insoluble in buffers. Under normal conditions, those proteins are either expressed at low levels, or/and will be removed by degradation machineries such as ubiquitin–proteasome pathway (UPP) and autophagosome–lysosome pathway (ALP). However, triggered by some environmental, pathological or/and aging factors, these proteins might be overexpressed, or/and the degradation machineries get inhibited. As a consequence, the proteins will accumulate and attack membranes to initiate various diseases including aging. Indeed, it has been recently revealed that immediately after synthesis, ~1–2% nascent proteins get degraded in yeast84 while the percentage of proteins that immediately got degraded can dramatically reach ~30% in humans85. Therefore, the homeostasis of various aggregated proteins in vivo appears to be the central factor responsible for a variety of human diseases including aging. The number and degree of the membrane attacking by aggregated proteins may serve as an endogenous clock to count down the aging process. Consequently, a key approach to fight against them is to develop strategies and agents to maintain or/and even enhance the functions of the degradation machineries22.\n\nThe human genomes appear to contain many pre-existing aggregated proteins like VAPB3. On the other hand, some family members or individuals carry additional genetic mutants like the P56S-VAPB which are insoluble in buffers. Under some environmental, pathological or/and aging conditions, these proteins might be overexpressed, or/and their degradation gets inhibited. As a consequence, the proteins will accumulate and attack membranes to initiate various diseases including aging.\n\n\nMaterials and methods\n\nThe expression and purification of the P56S-MSP domain followed the procedure we reported previously26. Briefly, the expression vectors were transformed into and overexpressed in Escherichia coli BL21 (DE3) cells (Novagen). The P56S-MSP protein was only found in inclusion bodies and consequently the pellet was first dissolved in a phosphate buffer (pH 8.5) containing 8 M urea and subsequently purified by a Ni2+-affinity column (Novagen) under denaturing conditions in the presence of 8 M urea. Dithiothreitol (DTT) was then added to the eluted fractions containing P56S-MSP to a final concentration of 100 mM to ensure complete conversion to Cys-SH. After 1 hour, the fractions were acidified by adding 10% acetic acid and subsequently purified by reverse-phase HPLC on a C4 column, and lyophilized.\n\nThe generation of the isotope-labeled proteins for NMR studies followed a similar procedure except that the bacteria were grown in M9 medium with the addition of (15NH4)2SO4 for 15N labeling and (15NH4)2SO4/[13C]-glucose for 15N-/13C-double labeling26. The purity of the recombinant proteins was checked by SDS–PAGE gels and their molecular weights were verified by a Voyager STR matrix-assisted laser desorption ionization time-of-flight-mass spectrometer (Applied Biosystems). The concentration of protein samples was determined by the UV spectroscopic method in the presence of 8 M urea86.\n\nThe P56S-MSP domain contains three free Cys residues at positions 41, 53 and 121. As such, the three Cys residues were first mutated to Ala by use of the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). Starting from this plasmid, a total of seven single-Cys mutants was prepared: Q6C, D24C, A53C, N68C, M89C, M102C and A121C (Figure 2B). The mutated plasmids were confirmed by DNA sequencing and their recombinant proteins were subsequently expressed and purified by the same procedures described above. 1H-15N heteronuclear single quantum coherence spectroscopy (HSQC) experiments were performed on each mutant to validate that these mutations did not significantly perturb the native structure of the P56S-MSP domain.\n\nThe recombinant proteins of seven single-cysteine mutants were Cys-modified following the previous procedure33–35, by the thiol-reactive nitroxide free radical probe, MTSSL (1-oxyl-2,2,5,5-tetramethyl-∆3-pyrroline-3-methyl) methanethiosulfonate (Toronto Research Chemicals Inc.). Briefly, the HPLC-purified recombinant protein of the each mutant was dissolved in the buffer containing 8 M urea, 20 mM phosphate (pH 8.0), which was pre-degassed with nitrogen gas for 20 minutes. Subsequently, the MTSSL reagent was added from 3.8 mM stock solution in acetonitrile to reach a ten-fold molar concentration of the protein, followed by incubation at room temperature with constant stirring for 5 hours. To ensure a complete labeling, another dose of MTSSL was added to a ten-fold molar concentration of the protein for an overnight incubation. The MTSSL-labeled protein was purified by reverse-phase HPLC on a C4 column and lyophilized. Based on the verification by the time-of-flight-mass spectrometer, the purity of the MTSSL-modified proteins of all mutants was >99% after the HPLC purification.\n\nAll circular dichroism (CD) experiments were performed on a Jasco J-810 spectropolarimeter equipped with a thermal controller using 1-mm path length cuvettes. Data from five independent scans were added and averaged26. The P56S-MSP samples were prepared at a protein concentration of 20 µM in either DMPC vesicles, bicelles formed by DMPC and DHPC, or DPC micelles in water (pH 4.0) and 5 mM phosphate (pH 7.5) respectively.\n\nAll NMR experiments were acquired on an 800 MHz Bruker Avance spectrometer equipped with pulse field gradient units as described previously47,48. NMR data were processed with NMRPipe87 and analysed with NMRView88. For characterizing the conformation of the P56S-MSP in water, a pair of triple-resonance experiments HNCACB, CBCA(CO)NH were collected for the sequential assignment on a 15N-/13C-double labelled sample in 90% H2O/10% D2O (pH 4.0). For achieving assignments of the P56S MSP domain in DPC micelles, triple-resonance experiments HNCACB, CBCA(CO)NH, HNCO, (H)CC(CO)NH, H(CCO)NH and HCCH-TOCSY were acquired on 15N-/13C-double labelled samples at a protein concentration of 500 µM in protonated DPC micelles (H-DPC) at 100 mM. For obtaining NOE connectivities, 13C-edited NOESY experiments were acquired on a double labeled P56S MSP sample in both H-DPC and deuterated DPC (D-DPC) micelles at 100 mM in D2O, while 15N-edited HSQC-TOCSY and HSQC-NOESY were collected on a 15N-labelled sample at a protein concentration of 500 µM in both H-DPC and D-DPC micelles at 100 mM in 90% H2O/10% D2O.\n\nFor assessing the backbone dynamics on the ps-ns time scale, {1H}-15N steady-state NOEs were obtained by recording spectra on the 15N-labeled P56S MSP domain at 500 µM in either water or H-DPC micelle (100 mM), with and without 1H presaturation with duration of 3 s plus a relaxation delay of 6 s at 800 MHz. To assess conformational exchanges over µs-ms, 15N transverse relaxation dispersion experiments were acquired on the P56S-MSP domain in H-DPC micelle, on a Bruker Avance 800 spectrometer with a constant time delay (TCP = 50 ms) and a series of CPMG frequencies, ranging from 40 Hz, 80 Hz, 120 Hz (x2), 160 Hz, 200 Hz, 240 Hz, 320 Hz, 400 Hz, 480 Hz, 560 Hz, 640 Hz, 720 Hz, 800 Hz, and 960 Hz (×2 indicates repetition) as we previously performed47,48. A reference spectrum without the CPMG block was acquired to calculate the effective transverse relaxation rate by the following equation:\n\n\n\nWhere I(νCPMG) is the peak intensity on the difference CPMG frequency and I0 is the peak intensity in the reference spectrum.\n\nTo probe the orientation of the P56S MSP residues, HSQC spectra of the P56S-MSP in H-DPC were acquired by gradual addition to 10 mM of manganese chloride and gadodiamide (gadolinium(III) 5,8-bis(carboxylatomethyl)-2-[2-(methylamino)-2-oxoethyl]-10-oxo-2,5,8,11-tetraazadodecane-1-carboxylate hydrate)35.\n\nBackbone dihedral angles were generated with TALOS+ by inputting backbone 1H, 15N and 13C chemical shifts38. NOE-based distance constraints were extracted from both 15N- and 13C-edited NOESY spectra collected on the P56S-MSP samples in D-DPC.\n\nParamagnetic relaxation enhancement (PRE) experiments were utilized to obtain long-range distance restraints. Specifically, for each spin-labeled single-cysteine mutant, a pair of 2D1H-15N HSQC spectra were acquired at a protein concentration of 200 µM in 40 mM H-DPC: one for the spin-labeled sample in the paramagnetic form, and another after adding ascorbic acid (to 10 mM) to the sample to reduce the nitroxide, yielding the diamagnetic sample. We also acquired HSQC spectra for 7 corresponding cysteine mutants without spin-labelling at the same conditions and only several HSQC peaks shifted after spin-labeling, indicating that the spin-labeling would not significantly change the conformation. The spectra were subsequently analyzed to obtain PRE-based differences in peak intensities using the programs nmrPipe86 as exemplified by Figure 8 showing the superimposition of HSQC spectra of the M89C mutant in the paramagnetic and diamagnetic forms.\n\nA. Overlay of two NMR HSQC spectra of spin-labeled M89C in the paramagnetic state of the MTSL probe (red) and diamagnetic state after the MTSL probe has been reduced (blue). The HSQC spectra were recorded on a 200 μM 15N-labeled M89C sample at 800 MHz and 313 K. B. Intensity ratios of HSQC peaks of spin-labeled M89C from the paramagnetic and diamagnetic states.\n\nIntensity ratios of peaks from the oxidized and reduced spectra were converted into PREs R2sp by estimating the additional transverse relaxation needed to reduce peak intensity relative to diamagnetic conditions by the observed intensity ratio as previously described33–35. Peaks unaffected by the paramagnetic probe (intensity ratio > 0.85) were not restrained while peaks with intensity ratio < 0.85 were converted to distances as previously described33–35. Structure calculations were carried out with the ab initio simulated annealing protocol of the Xplor-NIH program and CNS36,37. The NMR structures of the 125-residue P56S MSP domain in the DPC micelle have been deposited in PDB with ID of 2MDK. Protein structures were analyzed using PROCHECK89 and displayed by PyMol molecular graphics system (W. L. DeLano, DeLano Scientific LLC, San Carlos, CA).", "appendix": "Author contributions\n\n\n\nJXS conceived and designed the experiments. HNQ, JXS, LZL, YYW, GG performed the experiments and analyzed the data: JXS wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study is supported by Ministry of Education of Singapore (MOE) Tier 2 Grant MOE 2011-T2-1-096 (R154-000-525-112) to Jianxing Song.\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 Mr. Hong Meng Yew for preparing some protein samples.\n\n\nReferences\n\nRoss CA, Poirier MA: Protein aggregation and neurodegenerative disease. Nat Med. 2004; (Suppl): S10–7. PubMed Abstract \| Publisher Full Text\n\nChiti F, Dobson CM: Protein misfolding, functional amyloid, and human disease. Annu Rev Biochem. 2006; 75: 333–66. PubMed Abstract \| Publisher Full Text\n\nLindner AB, Demarez A: Protein aggregation as a paradigm of aging. Biochim Biophys Acta. 2009; 1790(10): 980–96. PubMed Abstract \| Publisher Full Text\n\nTannous P, Zhu H, Nemchenko A, et al.: Intracellular protein aggregation is a proximal trigger of cardiomyocyte autophagy. Circulation. 2008; 117(24): 3070–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNanga RP, Brender JR, Vivekanandan S, et al.: Structure and membrane orientation of IAPP in its natively amidated form at physiological pH in a membrane environment. Biochim Biophys Acta. 2011; 1808(10): 2337–2342. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu M, Hodish I, Haataja L, et al.: Proinsulin misfolding and diabetes: mutant INS gene-induced diabetes of youth. Trends Endocrinol Metab. 2010; 21(11): 652–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrender JR, Salamekh S, Ramamoorthy A: Membrane disruption and early events in the aggregation of the diabetes related peptide IAPP from a molecular perspective. Acc Chem Res. 2012; 45(3): 454–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZoghbi HY, Orr HT: Glutamine repeats and neurodegeneration. Annu Rev Neurosci. 2000; 23: 217–47. PubMed Abstract \| Publisher Full Text\n\nHolmes C, Boche D, Wilkinson D, et al.: Long-term effects of Abeta42 immunisation in Alzheimer's disease: follow-up of a randomised, placebo-controlled phase I trial. Lancet. 2008; 372(9634): 216–23. PubMed Abstract \| Publisher Full Text\n\nBruijn LI, Miller TM, Cleveland DW: Unraveling the mechanisms involved in motor neuron degeneration in ALS. Annu Rev Neurosci. 2004; 27: 723–49. PubMed Abstract \| Publisher Full Text\n\nNishimura AL, Mitne-Neto M, Silva HC, et al.: A mutation in the vesicle-trafficking protein VAPB causes late-onset spinal muscular atrophy and amyotrophic lateral sclerosis. Am J Hum Genet. 2004; 75(5): 822–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShi J, Lua S, Tong JS, et al.: Elimination of the native structure and solubility of the hVAPB MSP domain by the Pro56Ser mutation that causes amyotrophic lateral sclerosis. Biochemistry. 2010; 49(18): 3887–97. PubMed Abstract \| Publisher Full Text\n\nTsuda H, Han SM, Yang Y, et al.: The amyotrophic lateral sclerosis 8 protein VAPB is cleaved, secreted, and acts as a ligand for Eph receptors. Cell. 2008; 133(6): 963–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLua S, Qin H, Lim L, et al.: Structural, stability, dynamic and binding properties of the ALS-causing T46I mutant of the hVAPB MSP domain as revealed by NMR and MD simulations. PLoS One. 2011; 6(11): e27072. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVan Hoecke A, Schoonaert L, Lemmens R, et al.: EPHA4 is a disease modifier of amyotrophic lateral sclerosis in animal models and in humans. Nat Med. 2012; 18(9): 1418–22. PubMed Abstract \| Publisher Full Text\n\nNoberini R, Koolpe M, Peddibhotla S, et al.: Small molecules can selectively inhibit ephrin binding to the EphA4 and EphA2 receptors. J Biol Chem. 2008; 283(43): 29461–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQin H, Shi J, Noberini R, et al.: Crystal structure and NMR binding reveal that two small molecule antagonists target the high affinity ephrin-binding channel of the EphA4 receptor. J Biol Chem. 2008; 283(43): 29473–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTeuling E, Ahmed S, Haasdijk E, et al.: Motor neuron disease-associated mutant vesicle-associated membrane protein-associated protein (VAP) B recruits wild-type VAPs into endoplasmic reticulum-derived tubular aggregates. J Neurosci. 2007; 27(36): 9801–15. PubMed Abstract \| Publisher Full Text\n\nLi M, Liu J, Ran X, et al.: Resurrecting abandoned proteins with pure water: CD and NMR studies of protein fragments solubilized in salt-free water. Biophys J. 2006; 91(11): 4201–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu J, Song J: Insights into protein aggregation by NMR characterization of insoluble SH3 mutants solubilized in salt-free water. PLoS One. 2009; 4(11): e7805. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong J: Insight into \"insoluble proteins\" with pure water. FEBS Lett. 2009; 583(6): 953–9. PubMed Abstract \| Publisher Full Text\n\nSong J: Why do proteins aggregate? \"Intrinsically insoluble proteins\" and \"dark mediators\" revealed by studies on \"insoluble proteins\" solubilized in pure water [v1; ref status: indexed, http://f1000r.es/z0]. F1000Research. 2013; 2: 94. Publisher Full Text\n\nMitne-Neto M, Machado-Costa M, Marchetto MC, et al.: Downregulation of VAPB expression in motor neurons derived from induced pluripotent stem cells of ALS8 patients. Hum Mol Genet. 2011; 20(18): 3642–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFasana E, Fossati M, Ruggiano A, et al.: A VAPB mutant linked to amyotrophic lateral sclerosis generates a novel form of organized smooth endoplasmic reticulum. FASEB J. 2010; 24(5): 1419–30. PubMed Abstract \| Publisher Full Text\n\nPapiani G, Ruggiano A, Fossati M, et al.: Restructured endoplasmic reticulum generated by mutant amyotrophic lateral sclerosis-linked VAPB is cleared by the proteasome. J Cell Sci. 2012; 125(Pt 15): 3601–11. PubMed Abstract \| Publisher Full Text\n\nQin H, Wang W, Song J: ALS-causing P56S mutation and splicing variation on the hVAPB MSP domain transform its β-sandwich fold into lipid-interacting helical conformations. Biochem Biophys Res Commun. 2013; 431(3): 398–403. PubMed Abstract \| Publisher Full Text\n\nDelak K, Harcup C, Lakshminarayanan R, et al.: The tooth enamel protein, porcine amelogenin, is an intrinsically disordered protein with an extended molecular configuration in the monomeric form. Biochemistry. 2009; 48(10): 2272–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAguado-Llera D, Goormaghtigh E, de Geest N, et al.: The basic helix-loop-helix region of human neurogenin 1 is a monomeric natively unfolded protein which forms a \"fuzzy\" complex upon DNA binding. Biochemistry. 2010; 49(8): 1577–89. PubMed Abstract \| Publisher Full Text\n\nParker W, Sood A, Song A: Organization of regions with amphiphilic alpha-helical potential within the three-dimensional structure of beta-sheet proteins. Protein Eng. 2001; 14(5): 315–9. PubMed Abstract \| Publisher Full Text\n\nLee SY, Parker W: Amphiphilic α-helical potential: a putative folding motif adding few constraints to protein evolution. J Mol Evol. 2011; 73(3–4): 166–80. PubMed Abstract \| Publisher Full Text\n\nSchwarzinger S, Kroon GJ, Foss TR, et al.: Sequence-dependent correction of random coil NMR chemical shifts. J Am Chem Soc. 2001; 123(13): 2970–8. PubMed Abstract \| Publisher Full Text\n\nDyson HJ, Wright PE: Unfolded proteins and protein folding studied by NMR. Chem Rev. 2004; 104(8): 3607–22. PubMed Abstract \| Publisher Full Text\n\nBattiste JL, Wagner G: Utilization of site-directed spin labeling and high-resolution heteronuclear nuclear magnetic resonance for global fold determination of large proteins with limited nuclear overhauser effect data. Biochemistry. 2000; 39(18): 5355–65. PubMed Abstract \| Publisher Full Text\n\nBarrett PJ, Song Y, Van Horn WD, et al.: The amyloid precursor protein has a flexible transmembrane domain and binds cholesterol. Science. 2012; 336(6085): 1168–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVasudevan SV, Schulz J, Zhou C, et al.: Protein folding at the membrane interface, the structure of Nogo-66 requires interactions with a phosphocholine surface. Proc Natl Acad Sci U S A. 2010; 107(15): 6847–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrünger AT, Adams PD, Clore GM, et al.: Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54(Pt 5): 905–921. PubMed Abstract \| Publisher Full Text\n\nSchwieters CD, Kuszewski JJ, Tjandra N, et al.: The Xplor-NIH NMR molecular structure determination package. J Magn Reson. 2003; 160(1): 65–73. PubMed Abstract \| Publisher Full Text\n\nShen Y, Delaglio F, Cornilescu G, et al.: TALOS+: A hybrid method for predicting protein backbone torsion angles from NMR chemical shifts. J Biomol NMR. 2009; 44(4): 213–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLadokhin AS, White SH: Folding of amphipathic alpha-helices on membranes: energetics of helix formation by melittin. J Mol Biol. 1999; 285(4): 1363–9. PubMed Abstract \| Publisher Full Text\n\nWhite SH, Wimley WC: Membrane protein folding and stability: physical principles. Annu Rev Biophys Biomol Struct. 1999; 28: 319–65. PubMed Abstract \| Publisher Full Text\n\nFernández-Vidal M, Jayasinghe S, Ladokhin AS, et al.: Folding amphipathic helices into membranes: amphiphilicity trumps hydrophobicity. J Mol Biol. 2007; 370(3): 459–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlmeida PF, Ladokhin AS, White SH: Hydrogen-bond energetics drive helix formation in membrane interfaces. Biochim Biophys Acta. 2012; 1818(2): 178–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEngelman DM, Chen Y, Chin CN, et al.: Membrane protein folding: beyond the two stage model. FEBS Lett. 2003; 555(1): 122–5. PubMed Abstract \| Publisher Full Text\n\nTulumello DV, Deber CM: Efficiency of detergents at maintaining membrane protein structures in their biologically relevant forms. Biochim Biophys Acta. 2012; 1818(5): 1351–8. PubMed Abstract \| Publisher Full Text\n\nBowie JU: Structural biology. Membrane protein twists and turns. Science. 2013; 339(6118): 398–9. PubMed Abstract \| Publisher Full Text\n\nKay LE, Torchia DA, Bax A: Backbone dynamics of proteins as studied by 15N inverse detected heteronuclear NMR spectroscopy: application to staphylococcal nuclease. Biochemistry. 1989; 28(23): 8972–9. PubMed Abstract \| Publisher Full Text\n\nRan X, Qin H, Liu J, et al.: NMR structure and dynamics of human ephrin-B2 ectodomain: the functionally critical C-D and G-H loops are highly dynamic in solution. Proteins. 2008; 72(3): 1019–29. PubMed Abstract \| Publisher Full Text\n\nHuan X, Shi J, Lim L, et al.: Unique Structure and Dynamics of the EphA5 Binding Specificity Domain Mediate its Binding Specificity as Revealed by X-ray Crystallography, NMR and MD Simulations. PLoS ONE. 2013; 8(9): e74040. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTraaseth NJ, Veglia G: Probing excited states and activation energy for the integral membrane protein phospholamban by NMR CPMG relaxation dispersion experiments. Biochim Biophys Acta. 2010; 1798(2): 77–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLuckey M: Membrane Structural Biology: With Biochemical and Biophysical Foundations. Cambridge University Press: New York. 2008. Reference Source\n\nArinaminpathy Y, Khurana E, Engelman DM, et al.: Computational analysis of membrane proteins: the largest class of drug targets. Drug Discov Today. 2009; 14(23–24): 1130–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCramer WA, Engelman DM, Von Heijne G, et al.: Forces involved in the assembly and stabilization of membrane proteins. FASEB J. 1992; 6(15): 3397–402. PubMed Abstract\n\nDeber CM, Goto NK: Folding proteins into membranes. Nat Struct Biol. 1996; 3: 815–8. Reference Source\n\nWallin E, von Heijne G: Genome-wide analysis of integral membrane proteins from eubacterial, archaean, and eukaryotic organisms. Protein Sci. 1998; 7(4): 1029–1038. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKyte J, Doolittle RF: A simple method for displaying the hydropathic character of a protein. J Mol Biol. 1982; 157(1): 105–32. PubMed Abstract \| Publisher Full Text\n\nZhao G, London E: An amino acid \"transmembrane tendency\" scale that approaches the theoretical limit to accuracy for prediction of transmembrane helices: relationship to biological hydrophobicity. Protein Sci. 2006; 15(8): 1987–2001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEisenberg D, Weiss RM, Terwilliger TC: The hydrophobic moment detects periodicity in protein hydrophobicity. Proc Natl Acad Sci U S A. 1984; 81(1): 140–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrown LR, Lauterwein J, Wüthrich K: High-resolution 1H-NMR studies of self-aggregation of melittin in aqueous solution. Biochim Biophys Acta. 1980; 622(2): 231–44. PubMed Abstract \| Publisher Full Text\n\nLauterwein J, Brown LR, Wüthrich K: High-resolution. 1H-NMR studies of monomeric melittin in aqueous solution. Biochim Biophys Acta. 1980; 622(2): 219–30. PubMed Abstract \| Publisher Full Text\n\nVogel H: Incorporation of melittin into phosphatidylcholine bilayers: study of binding and conformationalchanges. FEBS Letters. 1981; 134(1): 37–42. PubMed Abstract \| Publisher Full Text\n\nChartier-Harlin MC, Kachergus J, Roumier C, et al.: Alpha-synuclein locus duplication as a cause of familial Parkinson's disease. Lancet. 2004; 364(9440): 1167–9. PubMed Abstract \| Publisher Full Text\n\nUlmer TS, Bax A, Cole NB, et al.: Structure and dynamics of micelle-bound human alpha-synuclein. J Biol Chem. 2005; 280(10): 9595–603. PubMed Abstract \| Publisher Full Text\n\nDevine MJ, Gwinn K, Singleton A, et al.: Parkinson's disease and α-synuclein expression. Mov Disord. 2011; 26(12): 2160–8. PubMed Abstract \| Publisher Full Text\n\nDikiy I, Eliezer D: Folding and misfolding of alpha-synuclein on membranes. Biochim Biophys Acta. 2012; 1818(4): 1013–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLashuel HA, Overk CR, Oueslati A, et al.: The many faces of α-synuclein: from structure and toxicity to therapeutic target. Nat Rev Neurosci. 2013; 14(1): 38–48. PubMed Abstract \| Publisher Full Text\n\nHedin LE, Illergård K, Elofsson A: An introduction to membrane proteins. J Proteome Res. 2011; 10(8): 3324–31. PubMed Abstract \| Publisher Full Text\n\nStevens TJ, Arkin IT: Turning an opinion inside-out: Rees and Eisenberg's commentary (Proteins 2000; 38: 121–122) on \"Are membrane proteins 'inside-out' proteins?\" (Proteins 1999; 36: 135–143). Proteins. 2000; 40(3): 463–4. PubMed Abstract \| Publisher Full Text\n\nAdamian L, Liang J: Prediction of transmembrane helix orientation in polytopic membrane proteins. BMC Struct Biol. 2006; 6: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhandelia H, Ipsen JH, Mouritsen OG: The impact of peptides on lipid membranes. Biochim Biophys Acta. 2008; 1778(7–8): 1528–36. PubMed Abstract \| Publisher Full Text\n\nBiverståhl H, Andersson A, Gräslund A, et al.: NMR solution structure and membrane interaction of the N-terminal sequence (1–30) of the bovine prion protein. Biochemistry. 2004; 43(47): 14940–7. PubMed Abstract \| Publisher Full Text\n\nElfrink K, Ollesch J, Stöhr J, et al.: Structural changes of membrane-anchored native PrP(C). Proc Natl Acad Sci U S A. 2008; 105(31): 10815–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShao H, Jao S, Ma K, et al.: Solution structures of micelle-bound amyloid beta-(1–40) and beta-(1–42) peptides of Alzheimer's disease. J Mol Biol. 1999; 285(2): 755–73. PubMed Abstract \| Publisher Full Text\n\nPoojari C, Kukol A, Strodel B: How the amyloid-β peptide and membranes affect each other: an extensive simulation study. Biochim Biophys Acta. 2013; 1828(2): 327–39. PubMed Abstract \| Publisher Full Text\n\nKünze G, Barré P, Scheidt HA, et al.: Binding of the three-repeat domain of tau to phospholipid membranes induces an aggregated-like state of the protein. Biochim Biophys Acta. 2012; 1818(9): 2302–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKegel KB, Sapp E, Yoder J, et al.: Huntingtin associates with acidic phospholipids at the plasma membrane. J Biol Chem. 2005; 280(43): 36464–73. PubMed Abstract \| Publisher Full Text\n\nSivanandam VN, Jayaraman M, Hoop CL, et al.: The aggregation-enhancing huntingtin N-terminus is helical in amyloid fibrils. J Am Chem Soc. 2011; 133(12): 4558–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMichalek M, Salnikov ES, Werten S, et al.: Membrane interactions of the amphipathic amino terminus of huntingtin. Biochemistry. 2013; 52(5): 847–58. PubMed Abstract \| Publisher Full Text\n\nKanekura K, Suzuki H, Aiso S, et al.: ER stress and unfolded protein response in amyotrophic lateral sclerosis. Mol Neurobiol. 2009; 39(2): 81–9. PubMed Abstract \| Publisher Full Text\n\nPfefferkorn CM, Jiang Z, Lee JC: Biophysics of α-synuclein membrane interactions. Biochim Biophys Acta. 2012; 1818(2): 162–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTosatto L, Andrighetti AO, Plotegher N, et al.: Alpha-synuclein pore forming activity upon membrane association. Biochim Biophys Acta. 2012; 1818(11): 2876–83. PubMed Abstract \| Publisher Full Text\n\nJang H, Connelly L, Arce FT, et al.: Mechanisms for the Insertion of Toxic, Fibril-like β-Amyloid Oligomers into the Membrane. J Chem Theory Comput. 2013; 9(1): 822–833. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpalding KL, Bhardwaj RD, Buchholz BA, et al.: Retrospective birth dating of cells in humans. Cell. 2005; 122(1): 133–43. PubMed Abstract \| Publisher Full Text\n\nNachreiner T, Esser M, Tenten V, et al.: Novel splice variants of the amyotrophic lateral sclerosis-associated gene VAPB expressed in human tissues. Biochem Biophys Res Commun. 2010; 394(3): 703–8. PubMed Abstract \| Publisher Full Text\n\nDuttler S, Pechmann S, Frydman J: Principles of cotranslational ubiquitination and quality control at the ribosome. Mol Cell. 2013; 50(3): 379–93. PubMed Abstract \| Publisher Full Text\n\nSchubert U, Antón LC, Gibbs J, et al.: Rapid degradation of a large fraction of newly synthesized proteins by proteasomes. Nature. 2000; 404(6779): 770–4. PubMed Abstract \| Publisher Full Text\n\nPace CN, Vajdos F, Fee L, et al.: How to measure and predict the molar absorption coefficient of a protein. Protein Sci. 1995; 4(11): 2411–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelaglio F, Grzesiek S, Vuister GW, et al.: NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR. 1995; 6(3): 277–93. PubMed Abstract \| Publisher Full Text\n\nJohnson BA, Blevins RA: NMRView: A computer program for the visualization and analysis of NMR data. J. Biomol. NMR. 1994; 4(5): 603–614. PubMed Abstract \| Publisher Full Text\n\nLaskowski R, MacArthur MW, Moss DS, et al.: PROCHECK: A program to check the sterochemical quality of protein structures. J Appl Crystallogr. 1993; 26: 283–291. Reference Source" }	[ { "id": "4030", "date": "21 Mar 2014", "name": "Annalisa Pastore", "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 Qin et al. describes a structural study of the MSP domain of VAP8 in a membrane-like environment (DPC micelles). The work is overall interesting. However, I have a number of issues that I would like the authors to consider:The title is too general. I do not think that such a generalization is possible at this stage. I think it should be changed into something more relevant for the specific protein. I needed to read down in the introduction to understand which protein it refers to. The VAO8 protein must be cited already in the abstract (and possibly in the title). Some more details on the domain must be given. Is this a beta sandwich? It looks like anIg-like fold. It should be commented on. I do not think the first sentence of the abstract is justified at this level of generalization. I would rewrite it. I am not even sure it is correct. Do the authors refer to the well known problem that toxicity is observed also in the absence of detectable aggregates? It should be clarified. Some of the figures are unnecessary. I would delete figure 5. Figure 6 is unnecessarily large. Most of the figure is unsupported by the data shown in the manuscript. It could be simplified without altering its meaning. The authors should avoid sentences such as: ‘Unexpectedly, five well-formed helices in the membrane environment adopted β-strands in the native MSP fold…’. Why unexpectedly? It is well known that many sequences become helical when in micelles. This has been known for 30 years (my PhD thesis was on a structural study of a peptide in SDS). Likewise, the authors seem to be surprised by observations that have already been reported by others. Without wanting to take anything away from the authors, the idea of a role of poration in neurodegeneration is rather old and has been suggested for Abeta, polyglutamines and several other proteins. Along the logic of 4., the authors should really cite Kagan et al. (2012) and Crescenzi et al. (2002).Overall the work is fine and solid but the authors should turn down any claim of generality and/or absolute novelty. The discussion is clear.", "responses": [] }, { "id": "2268", "date": "15 Jul 2014", "name": "H Jane Dyson", "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 seems like a very interesting new idea, and the experiments seem well done.I do find the title a bit confusing. if the proteins are aggregated, how is it that they attack the membranes without needing aggregates? Maybe a qualification such as \"a common mechanism for aggregation-prone proteins to attack membranes without the participation of large aggregates\"?The only other small change I would recommend is to check the title of Figure 2. Should this be NMR not NME?", "responses": [] } ]	1	https://f1000research.com/articles/2-221
https://f1000research.com/articles/3-166/v1	22 Jul 14	{ "type": "Case Report", "title": "Case Report: Pneumocephalus after labor epidural anesthesia", "authors": [ "Beatriz Nistal-Nuño", "Manuel Ángel Gómez-Ríos ", "Manuel Ángel Gómez-Ríos " ], "abstract": "Lumbar epidural anesthesia is commonly used for labor analgesia. The 'loss-of- resistance' to air technique (LORA) is generally employed for recognition of the epidural space. One of the rare complications of this technique is pneumocephalus (PC). Here we describe the case of a parturient who developed a frontal headache when locating the epidural space using LORA. On the second day after epidural injection, the patient exhibited occipital headaches with gradual worsening. Computed tomography scans of the brain indicated PC. Following symptomatic treatment, our patient was discharged on the 13th day. We concluded that the amount of air used to identify the epidural space in LORA should be minimized, LORA should not be used after dural puncture and the use of saline avoids PC complications.", "keywords": [ "Epidural anesthesia", "complication", "pneumocephalus" ], "content": "Introduction\n\nLumbar epidural anesthesia is a commonly used technique for analgesia during labor. The complications often associated to this technique include unilateral analgesia, extended epidural blockade, unplanned puncture of the dura or of a blood vessel, post-puncture dural headache (PDPH), subdural blockade, placement of the catheter out of the epidural space and neurological complications1. The ‘loss-of-resistance’ to air technique (LORA) is commonly employed for recognition of the epidural space. Nevertheless, one of its rare complications is pneumocephalus (PC)2.\n\nCases of PC following neuroaxial anesthesia have been described either after an epidural using the LORA technique or after spinal anesthesia3,4. The development of PC after spinal anesthesia is exceptionally rare5,6. PC symptoms are difficult to distinguish from other complications of the epidural technique such as PDPH or neurotoxicity. Diagnosis relies on clinical impression and brain tomography (CT-scan)2. Our goal is to report the case of a patient who presented PC following labor epidural anesthesia.\n\nWe attempted to approach the patient and the patient’s family to obtain informed consent for publication of this report, However we decided to abandon telephone contact after numerous unsuccessful attempts. The Institutional Review Board at Complexo Hospitalario Universitario A Coruna determined that approval of the case report was not required.\n\n\nClinical case\n\nA 34 year old healthy Caucasian parturient ASA II, G1P0 was admitted at 38 weeks of gestation. Her clinical history included no allergies to medications, no relevant family history, mild bicuspid aortic valve stenosis, cervical aortic arch and mild postductal coarctation of the aorta. She was not taking any medications. She presented occasional asymptomatic palpitations at rest (few seconds long). Her cardiopathy was compensated and well controlled during pregnancy.\n\nAt 4–5 cm of cervical dilatation, at the request of analgesia, a lumbar epidural was proposed for management of labor pain. The patient was monitored with ECG, SpO2, non-invasive blood pressure (NIBP) every 5 minutes and a peripheral vessel catheter was in situ. Puncture was performed at the L4–L5 interspace with a 18 G Tuohy epidural needle (Perifix® 401 Braun Germany) (80 mm/3¼\" long) via the median approach with the patient in the sitting position. After locating the epidural space using the LORA technique (approximately 3 ml of air), the patient developed a sudden intense frontal headache. No neurological, haemodynamic changes or breathing symptoms were detected. The needle was withdrawn without cerebrospinal fluid (CSF) flashback. The cephalalgia improved gradually ceasing several minutes later. An epidural catheter was introduced afterwards (Perifix® 401 Braun Germany, close-end, three lateral holes) in the L3–L4 interspace with the patient in the sitting position, after locating the epidural space using loss of resistance to saline (LORS). We confirmed its appropriate location by checking that CSF had not been aspirated prior to local anesthetic injection. The test dose of 3 ml of bupivacaine 0.25% and epinephrine resulted negative. An initial bolus of 5 ml of bupivacaine 0.25% and 50 mcg of fentanyl was administered and a continuous infusion of ropivacaine 0.18% in addition to 1 mcg of fentanyl per ml was programmed at 7 ml/hour. This process was incident-free, being the epidural analgesia correct and labor with normal evolution.\n\nOn the second day postpartum, the patient presented occipital headaches which increased on standing, accompanied by tinnitus, nausea and vomiting, with gradual worsening. We found no alterations in the neurological examination. We decided conservative treatment with rest, NSAIDs, intravenous hyperhydration and 300 mg of oral caffeine twice daily, showing mild clinical improvement. However, on the 6th day after delivery, due to mild clinical worsening, and suspecting iatrogenic PC vs. PDPH, a brain CT-scan was performed. The CT scan revealed air in the temporal horns and right frontal horn of the lateral ventricle, showing a PC in the subarachnoid space and ventricles that was responsible for mild ventricular dilation (Figure 1 and Figure 2). A subdural collection that could correspond to an inflammatory reaction was also observed. At all times the neurological examination was normal. We consulted the Neurosurgery Service, which considered the PC not significant and recommended to continue with the conservative treatment and repeat the CT in 48–72 hours or if neurological changes occurred.\n\nDespite treatment, at the 8th day after puncture, the patient’s symptoms worsened at night with severe frontal headaches (Visual Analog Scale 6) and an increase in tinnitus. She was afebrile, the vital signs were unremarkable, and she had no focal neurological deficits. A control CT scan showed a mild regression of the PC in the ventricular system, but demonstrated persistence of the subdural collection. We decided to continue with the conservative treatment and performed control blood tests (routine hematological; liver and kidney function tests; serum electrolytes; coagulation profile).\n\nOn the 13th day after delivery, the follow-up was uneventful, the patient showed no abnormalities at the neurological examination and her vital signs were normal. The patient was finally discharged from the anesthesiology department being completely asymptomatic.\n\n\nDiscussion\n\nPC is relatively common in neurosurgery7 and neuroradiology8,9. It can be caused by trauma10 or infections11,12. It may develop after lumbar puncture, epidural steroid injection, or Valsalva’s maneuver13–15. The development of PC after spinal or epidural anesthesia is extremely infrequent. The incidence of PC after epidural steroid injections or epidural anesthesia is unknown, and only few cases per year are described in the literature16,17. Most cases of PC due to epidural techniques have been associated with LORA, as described in our case report.\n\nPC, an unusual consequence of evident or unnoticed accidental dural puncture3,4,18, develops from the injection of air into the subarachnoid or subdural space and cranial migration19. PC is not often followed by symptoms, but among those, headache is the most frequent20,21. The appearance of other symptoms, such as signs of space-occupying lesions (focal neurologic deficits including cranial nerve palsies19,20, or diverse motor signs) or augmented intracranial pressure and cardiovascular instability may develop depending on the spread and extent of intracranial air22. Headache is caused by the fast brain motion resulted from air injection and meningeal irritation21. Most cases consist of abrupt intense frontotemporal cephalea5,23,24, as in our case study, having a premature beginning (same day) and commonly concluding within 5 days. It is exacerbated by motion and may not be alleviated by lying down2. Roderick et al. outlined that 2 ml of air injected into the subarachnoid space was sufficient to provoke a symptomatic PC5.\n\nIn case of PDPH, due to CSF outflow, the pain is exacerbated by sitting or standing and is alleviated by lying down, having a characteristic occipital, frontal and post-orbital situation. It happens more often 24 to 48 hours after dural puncture25 and is longer lasting than in PC2. Although there may be subtle clinical differences with PC, their symptoms usually are interchangeable so the differential diagnosis must be done through CT.\n\nA number of techniques to find the epidural space have been defined26. LORA and LORS are the most common methods used2. Potential inconveniences of using saline comprise the difficulty to ascertain a meningeal puncture27. On the other hand, if air is forced quickly by digital pressure, false positives may result, or gas embolism, subcutaneous emphysema28 or multiradicular syndromes29.\n\nAccidental dural puncture is not always evident, as also shown in our case. Okell and Springge30 described a 0.6% incidence of dural punctures in epidural anesthesia. These punctures were acknowledged by the loss of CSF, by aspiration through the catheter, by hypotension after injection of a test dose, and retrospectively. Hardy31 reported that an epidural catheter cannot easily be passed through the dura, but the arachnoid can be penetrated smoothly. The author deduced that when a catheter goes into the subarachnoid space it is due to its initial subdural placement and movement to the arachnoid, as the first hole in the dura yields migration of the catheter2. Air introduced into the subdural space is more painful than in the subarachnoid space and reaches rapidly the head, because of its low pressure and diminished capacitance32,33.\n\nThese findings explain the sequence in our case. We found no evidence of dural puncture with CSF flashback with the needle insertion at the L4-L5 interspace, but the dura had already been penetrated, allowing the passage of air likely into the subdural space and causing the abrupt headache. Despite our uncertainty regarding dural puncture, we correctly avoided LORA and used LORS for the following attempt to localize the epidural space.\n\nIn our case, PC was diagnosed from the CT scan. While subdural space is not straightly connected to the subarachnoid space34, it is annexed to the floor of the third ventricle in the cranial cavity from the lower border of the second sacral vertebra35,36.\n\nThe treatment of PC consists on administration of 40–100% oxygen in the supine position37. This is to favor the reabsorption of intracranial air by intensifying the diffusion concentration gradient for nitrogen between the air collection and the surrounding cerebral tissue38. Nitrous oxide should be avoided to prevent the expansion of PC39. In addition, we should administer aggressive hydration, caffeine, or analgesics40,41. Epidural infusion or blood patch have no effect on PC3,19.\n\nThere is usually reabsorption of the air within 3–5 days from the epidural injection and patients commonly improve without any neurologic abnormalities. Our patient was discharged after clinical-radiologic resolution on the 13th day. If tension PC occurs, a neurosurgical emergency treatment may be necessary21,23,42.\n\nIn conclusion, the way to minimize the likelihood of PC when performing epidural block with the loss of resistance technique is to use saline instead of air43,44. When air is used, we should minimize its extent. In addition, LORA should not be used after dural puncture.\n\nAfter epidural block, and particularly when dural puncture is performed, the patient should be monitored carefully. We should identify that the headache from PC after epidural anesthesia occurs commonly immediately after puncture. Likewise, we should recognize that symptoms of PC are similar to those of PDPH and that a differential diagnosis is established by imaging techniques. Lastly, we should be able to assess that PC may be spontaneously absorbed, managed with symptomatic treatment.\n\n\nConsent\n\nAfter numerous unsuccessful attempts to contact the patient and the patient’s family to obtain informed written consent for publication of this report, the Institutional Review Board at our Institution determined that approval of the case report was not required.", "appendix": "Author contributions\n\n\n\nBeatriz Nistal-Nuño: writing and elaboration of the entire manuscript (all sections of the manuscript-abstract, introduction, case description, discussion, references, including its translation and editing), collection of data, and interpretation of the clinical case.\n\nManuel Gomez-Rios: collection of data.\n\nBoth authors revised 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\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nParnass SM, Schmidt KJ: Adverse effects of spinal and epidural anaesthesia. Drug Saf. 1990; 5(3): 179–194. PubMed Abstract \| Publisher Full Text\n\nMateo E, López-Alarcón MD, Moliner S, et al.: Epidural and subarachnoidal pneumocephalus after epidural technique. Eur J Anaesthesiol. 1999; 16(6): 413–7. PubMed Abstract \| Publisher Full Text\n\nAida S, Taga K, Yamakura T, et al.: Headache after attempted epidural block: the role of intrathecal air. Anesthesiology. 1998; 88(1): 76–81. PubMed Abstract\n\nAlonso E, Parajua JL: [Iatrogenic pneumocephalus caused by epidural analgesia]. Neurologia. 2000; 15(9): 418. PubMed Abstract\n\nRoderick L, Moore DC, Artru AA: Pneumocephalus with headache during spinal anesthesia. Anesthesiology. 1985; 62(5): 690–2. PubMed Abstract \| Publisher Full Text\n\nAvellanal M, Olmedilla L, Ojea R, et al.: Pneumocephalus after spinal anesthesia. Anesthesiology. 1996; 85(2): 423–5. PubMed Abstract\n\nReasoner DK, Todd MM, Scamman FL, et al.: The incidence of pneumocephalus after supratentorial craniotomy. Observations on the disappearance of intracranial air. Anesthesiology. 1994; 80(5): 1008–12. PubMed Abstract \| Publisher Full Text\n\nPeterson HO, Kieffer SA: Neuroradiology. Volume 1. New York, Hayon and Row. 1984; 127–30.\n\nHeinz ER: Techniques in imaging of the spine, part 3: Myelography, The Clinical Neurosciences: Neuroradiology. Edited by Rosenberg RN. New York, Churchill Livingstone. 1984; 795–817.\n\nFinch MD, Morgan GA: Traumatic pneumocephalus following head injury. A complication of general anaesthesia. Anaesthesia. 1991; 46(5): 385–7. PubMed Abstract \| Publisher Full Text\n\nCandan S, Katelioglu M, Ceylan S, et al.: Otogenic brain abscess with pneumocephalus. Infection. 1990; 18(3): 191–2. PubMed Abstract \| Publisher Full Text\n\nFinelli P, Bergen R: Spontaneous pneumocephalus with meningitis: CT demonstration. J Comput Assist Tomogr. 1991; 15(3): 525–6. PubMed Abstract\n\nOzturk E, Kantarci M, Karaman K, et al.: Diffuse pneumocephalus associated with infratentorial and supratentorial hemorrhages as a complication of spinal surgery. Acta Radiol. 2006; 47(5): 497–500. PubMed Abstract \| Publisher Full Text\n\nClevens RA, Marentette LJ, Esclamado RM, et al.: Incidence and management of tension pneumocephalus after anterior craniofacial resection: case reports and review of the literature. Otolaryngol Head Neck Surg. 1999; 120(4): 579–83. PubMed Abstract \| Publisher Full Text\n\nYoon SJ, Oh GS, Lee SJ, et al.: Pneumocephalus in patients with orthostatic headache. J Clin Neurol. 2008; 4(2): 89–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHawley JS, Ney JP, Swanberg MM: Subarachnoid pneumocephalus from epidural steroid injection. Headache. 2005; 45(3): 247–8. PubMed Abstract \| Publisher Full Text\n\nSimopoulos T, Peeters-Asdourian C: Pneumocephalus after cervical epidural steroid injection. Anesth Analg. 2001; 92(6): 1576–7. PubMed Abstract \| Publisher Full Text\n\nAhlering JR, Brodsky JB: Headache immediately following attempted epidural analgesia in obstetrics. Anesthesiology. 1980; 52(1): 100–101. PubMed Abstract\n\nNafiu OO, Urquhart JC: Pneumocephalus with headache complicating labour epidural analgesia: should we still be using air? Int J Obstet Anesth. 2006; 15(3): 237–9. PubMed Abstract \| Publisher Full Text\n\nKatz JA, Lukin R, Bridenbaugh PO, et al.: Subdural intracranial air: an unusual cause of headache after epidural steroid injection. Anesthesiology. 1991; 74(3): 615–618. PubMed Abstract \| Publisher Full Text\n\nKim YD, Lee JH, Cheong YK: Pneumocephalus in a patient with no cerebrospinal fluid leakage after lumbar epidural block - a case report -. Korean J Pain. 2012; 25(4): 262–266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGómez-Ríos MÁ, Fernández-Goti MC: Pneumocephalus after inadvertent dural puncture during epidural anesthesia. Anesthesiology. 2013; 118(2): 444. PubMed Abstract \| Publisher Full Text\n\nKatz Y, Morkovits R, Rosenberg B: Pneumoencephalos after inadvertent intrathecal air injection during epidural block. Anesthesiology. 1990; 73(6): 1277–9. PubMed Abstract \| Publisher Full Text\n\nGonzalez-Carrasco FJ, Aguilar JL, Llubia C, et al.: Pneumocephalus after accidental dural puncture during epidural anesthesia. Reg Anesth. 1993; 18(3): 193–5. PubMed Abstract\n\nBenzon HT, Linde HW, Molloy RE, et al.: Postdural puncture headache in patients with chronic pain. Anesth Analg. 1980; 59(10): 772–774. PubMed Abstract\n\nDe Andrés JA, Gomar C, Calatrava P, et al.: Comparative study of detection methods in epidural anesthesia: Episensor and loss of resistance. Rev Esp Anestesiol Reanim. 1990; 37(6): 330–334. PubMed Abstract\n\nValentine SJ, Jarvis AP, Shutt LE: Comparative study of the effects of air or saline to identify the extradural space. Br J Anaesthesia. 1991; 66(2): 224–227. PubMed Abstract \| Publisher Full Text\n\nViel EJ, de La Coussaye JE, Bruelle P, et al.: Epidural anesthesia: a pitfall due to the technique of the loss of resistance to air. Reg Anesth. 1991; 16(2): 117–119. PubMed Abstract\n\nRozenberg B, Tischler S, Glick A: Abdominal subcutaneous emphysema: an unusual complication of lumbar epidural block. Can J Anaesth. 1988; 35(3 (Pt 1)): 325. PubMed Abstract \| Publisher Full Text\n\nOkell RW, Sprigge JS: Unintentional dural puncture: a survey of recognition and management. Anaesthesia. 1987; 42(10): 1100–1111. PubMed Abstract\n\nHardy PA: Can epidural catheters penetrate dura mater? An anatomical study. Anaesthesia. 1986; 41(11): 1146–1147. PubMed Abstract \| Publisher Full Text\n\nHogan QH, Haddox JD: Headache from intracranial air after a lumbar epidural injection: subarachnoid or subdural? Reg Anesth. 1992; 17(5): 303–5. PubMed Abstract\n\nSmith GB, Barton F, Watt J: Extensive spread of local anesthetic solution following subdural insertion of an epidural catheter during labour. Anaesthesia. 1984; 39(4): 355–8. PubMed Abstract \| Publisher Full Text\n\nAjar AH, Rathmell JP, Mukherji SK: The subdural compartment. Reg Anesth Pain Med. 2002; 27(1): 72–6. PubMed Abstract \| Publisher Full Text\n\nCollier CB: Why obstetric epidurals fail: a study of epidurograms. Int J Obstet Anesth. 1996; 5(1): 19–31. PubMed Abstract \| Publisher Full Text\n\nReina MA, De Leon Casasola O, López A, et al.: The origin of the spinal subdural space: ultrastructure findings. Anesth Analg. 2002; 94(4): 991–5. PubMed Abstract \| Publisher Full Text\n\nKim YJ, Baik HJ, Kim JH, et al.: Pneumocephalus developed during epidural anesthesia for combined spinal-epidural anesthesia. Korean J Pain. 2009; 22(2): 163–166. Publisher Full Text\n\nDexter F, Reasoner DK: Theoretical assessment of normobaric oxygen therapy to treat pneumocephalus. Anesthesiology. 1996; 84(2): 442–447. PubMed Abstract \| Publisher Full Text\n\nMcMurtrie R Jr, Jan R: Subarachnoid pneumocephalus: a rare complication of epidural catheter placement. J Clin Anesth. 2002; 14(7): 539–42. PubMed Abstract \| Publisher Full Text\n\nReddy HV, Queen S, Prakash D, et al.: Tension pneumocephalus: an unusual complication after lung resection. Eur J Cardiothorac Surg. 2003; 24(1): 171–173. PubMed Abstract \| Publisher Full Text\n\nBilsky MH, Downey RJ, Kaplitt MG, et al.: Tension pneumocephalus resulting from iatrogenic subarachnoid-pleural fistulae: report of three cases. Ann Thorac Surg. 2001; 71(2): 455–457. PubMed Abstract \| Publisher Full Text\n\nBlack PM, Davis JM, Kjellberg RN, et al.: Tension pneumocephalus of the cranial subdural space: a case report. Neurosurgery. 1979; 5(3): 368–370. PubMed Abstract \| Publisher Full Text\n\nvan den Berg AA, Nguyen L, von-Maszewski M, et al.: Unexplained fitting in patients with post-dural puncture headache. Risk of iatrogenic pneumocephalus with air rationalizes use of loss of resistance to saline. Br J Anaesth. 2003; 90(6): 810–1. PubMed Abstract \| Publisher Full Text\n\nLee SS, Lee SG, Min BW, et al.: Pneumocephalus after an autologous blood patch procedure to treat a postdural puncture headache: A case report. Korean J Anesthesiol. 2008; 55(3): 387–390. Publisher Full Text" }	[ { "id": "5538", "date": "29 Jul 2014", "name": "Richard Smiley", "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 have reported an interesting case of headache after a lumbar epidural procedure. There is value in reminding anesthesiologists of the potential for pneumocephalus (PC) when air is used for the loss-of-resistance procedure. Their description makes it clear that the patient developed an almost immediate headache, which resolved rather quickly. This was followed by a more chronic, at least partially postural headache starting day 2 after the procedure. While the initial headache was almost certainly due to air injection into the subdural or subarachnoid space and a resulting PC, the subsequent headache may well have been a more classic post-dural puncture headache (PDPH), perhaps with some residual from the PC, given its time course (starting day 2, continuing for the better part of two weeks, which is later and longer than most PC headaches, but exactly what one might expect from a PDPH).\n\nThe authors mention that PC headaches typically last less than 5 days (the few we have seen have tended to last hours not days). A bit more discussion of this possibility/probability could be in the manuscript. It is possible, especially when one considers that the neurosurgeons considered the PC “”not significant,” that the patient could have been treated with a epidural blood patch to relieve her headache and shorten her hospital stay. In general, PC headaches are not as obviously postural as PDPH, so while certainly symptoms overlap, I am not sure it is fair for the authors to state that the symptoms are “interchangeable.” The authors suggest that because no CSF was seen with the initial placement the Tuohy needle may have been in the subdural space, and give an anatomic explanation of how air gets form the subdural to the intracerebral subarachnoid space; this is a possible explanation, but it is also possible that the air was injected subarachnoid but the needle withdrawn before CSF was appreciated. It is interesting that Roderick, in 1985 (ref 5 in this case report) reported a PC after injection of ~ 2ml air during attempted spinal anesthesia; in that case too there were problems obtaining CSF; perhaps they were subdural. There are some unimportant issues with English usage that do not detract from understanding the main points of the case report. I am not entirely sure I understand the mechanism of headache from PC as described in the case report. The authors cite “fast brain motion” and “meningeal irritation.” Any brain movement must be short-lived, explaining only the immediate, initial headache, and I am not sure how air is much of an “irritant.” Unless the air is of a very significant volume, it seems likely that it has quite small effects after a few hours or a day or two; hence my suspicion this case is more a PDPH than PC headache.", "responses": [] }, { "id": "5543", "date": "04 Aug 2014", "name": "Jean Charchaflieh", "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 provided an excellent description of a case of pneumocephalus (PC) after epidural catheterization, without evident dural puncture, with very informative discussion of the pathophysiology, symptomatology and management. The authors discussed briefly some of the advantages and disadvantages of using loss of resistance to air (LORA) versus loss of resistance to saline (LORS) in identifying the epidural space. The authors provided three recommendations for the prevention of PC after epidural catheterization, which are not necessarily supported by empirical evidence or consensus of practitioners. The three recommendations are:1. Use LORS instead of LORA to minimize the likelihood of PC after epidural catheterization.This recommendation is not supported by empiric evidence or consensus of practitioners. Indeed one of the cited references in support of this recommendation indicates that all three reported cases of PC occurred after LORS; and that the general practice in UK is about 50% LORS and 50% LORA.43 in main article As the authors noted, one of the potential disadvantages of LORS is clouding the certainty of the occurrence of dural puncture.27 in main articleAnother potential disadvantage of LORS is creating a pocket of fluid in front of ligamentum flavum that allows placing the epidural catheter in front of ligamentum flavum instead of the epidural space.The argument that LORS minimizes the likelihood of PC after epidural catherization is not clear cut and not the overriding consideration in choosing LORA vs. LORS.2. Minimize the amount of air when using LORA.When using LORA, most practitioners would argue in favor of limiting the amount of injected air to 1-2 ml, and against the use of multiple confirmatory injections of air into the epidural space. However as the presented case demonstrate, PC can occur even when performing LORA with approximately 3 ml of air and without evident dural puncture.3. LORA should not be used after dural puncture.This recommendation is not supported by empiric evidence and not universally endorsed by practitioners. As noted above, LORA can be more advantageous than LORS at detecting dural puncture. As for the immediate management of dural puncture, interventions can vary from inserting a catheter into the intrathecal space to reattempting the epidural catheterization at a different lumbar level. No specific technique has proven an advantage or disadvantage regarding the risk of PC. Indeed, as the presented case demonstrates, PC can occur even when LORS is used after initial LORA without excessive amount of air and without evident dural puncture. It is worth noting that post-dural puncture headache (PDPH) can co-exist and outlast headache due to PC. One of the referees (IV) reported a case of PC for 3 days, followed by resolution of PC and subsequent occurrence of typical PDPH, which resolved with treatment with epidural blood patch (Velickovic & Pavlik, 2007). Overall, the authors presented a well-balanced case report and discussion, which can serve as a handy review for diagnosis and management of PC after epidural catheterization.", "responses": [] }, { "id": "5559", "date": "14 Aug 2014", "name": "Jean Eloy", "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 case report: “Pneumocephalus after epidural anesthesia” is a well written manuscript. The Hypothesis is sound and the originality of the manuscript evident. The report is also relevant to clinical practice, shows appropriate methodology and is properly referenced. However, in the introduction the author states that the development of PC after spinal anesthesia is exceptionally rare - which it is not. I would omit the word exceptionally from the manuscript, as PC is often confused with post dural puncture headache and is often underreported.As the author mentioned in the conclusion section, using loss of resistance to saline instead of air is a way to minimize the incidence of PC. However the author does not expand on which technique is used the most by practitioners and taught in residency programs. Although this case report is hardly the definitive paper on Pneumocephalus after epidural analgesia, it certainly offers guidance and raises awareness on its incidence.From the case report, the author report sudden onset of frontal headache that improved within several minutes. There was no mention in the case report or the discussion on when the epidural catheter was removed and the temporal relationship with the patient occipital headaches. The signs and symptoms of PDPH and PC are the same. Are the symptoms reported by the\n\nauthor the results of PDPH or PC? This has not been made", "responses": [] } ]	1	https://f1000research.com/articles/3-166
https://f1000research.com/articles/2-238/v1	11 Nov 13	{ "type": "Research Article", "title": "Functional characterization of Candida albicans Hos2 histone deacetylase", "authors": [ "G Karthikeyan", "Maneesh Paul-Satyaseela", "Nachiappan Dhatchana Moorthy", "Radha Gopalaswamy", "Shridhar Narayanan", "Maneesh Paul-Satyaseela", "Nachiappan Dhatchana Moorthy", "Radha Gopalaswamy", "Shridhar Narayanan" ], "abstract": "Candida albicans is a mucosal commensal organism in normal individuals, but is a major pathogen causing systemic and mucosal infections in immunocompromised individuals. Azoles have been very effective anti-fungal agents and the mainstay in treating opportunistic mold and yeast infections. Azole resistant strains have emerged compromising the utility of this class of drugs. It has been shown that azole resistance can be reversed by the co-administration of a histone deacetylase (HDAC) inhibitor, suggesting that resistance is mediated by epigenetic mechanisms possibly involving Hos2, a fungal deacetylase. We report here the cloning and functional characterization of HOS2 (High Osmolarity Sensitive), a gene coding for fungal histone deacetylase from C. albicans. Inhibition studies showed that Hos2 is susceptible to pan inhibitors such as trichostatin A (TSA) and suberoylanilide hydroxamic acid (SAHA), but is not inhibited by class I inhibitors such MS-275. Purified Hos2 protein consistently deacetylated tubulins, rather than histones from TSA-treated cells. This in vitro enzymatic assay, which is amenable to high throughput could be used for screening potent fungal Hos2 inhibitors that could be a potential anti-fungal adjuvant. Hos2 has been reported to be a putative NAD+ dependent histone deacetylase, a feature of sirtuins. We assayed for sirtuin activation with resveratrol and purified Hos2 protein and did not find any sirtuin activity.", "keywords": [ "C. albicans", "HOS2", "enzyme assay", "histone/ tubulin deacetylase activity." ], "content": "Introduction\n\nCandida albicans is a commensal organism found in the mucosa and gastrointestinal tract of most healthy individuals but can cause severe systemic/superficial infections, especially in immunocompromised patients1. This fungus is of clinical importance and is one of the leading causes of systemic infections in immunocompromised individuals. C. albicans is the fourth most common cause of nosocomial bloodstream infections and is associated with high mortality rates2. Azoles and echinocandins targeting the ergosterol and cell wall biosynthesis pathway respectively have been used as anti-fungal drugs though the emergence of drug-resistant strains has compromised the efficacy and utility of these drugs3.\n\nAzole resistance in Candida sp. is mediated by up regulation of genes encoding ERG11, a lanosterol demethylase4–6, MDR15,7 and by CDR (Candida drug resistance) efflux pumps5,7–9. A combination of existing anti-fungals with new classes of drugs that act by different mechanisms will be viable alternatives to the current monotherapy regimen, which contributes to the emergence of drug resistance.\n\nHistone deacetylases (HDAC) play an important role in modulating chromatin conformation, by deacetylating crucial lysine residues in the histone octamers over which the chromatin DNA are wrapped10. Human HDAC`s fall into four broad categories, Class I (HDAC1, 2, 3, and 8), Class II a (HDAC 4, 5, 7 and 9) Class II b (HDAC 6 and 10) Class III (sirtuins) and Class IV (HDAC11) based on sequence homology, substrate preference and co-factor requirements. The involvement of each of these isoforms in disease pathology has been elucidated to some extent in recent times. The approval of suberoylanilide hydroxamic acid (SAHA)11, a well known inhibitor of HDACs by the US FDA for treating CTCL, (cutaneous T cell lymphoma)12,13 has thrown open the doors for exploring the use of HDAC inhibitors in combination with existing drugs for several diseases, such as malaria and Kala-azar etc.,14–16.\n\nClass specific inhibitors are now becoming a reality for human HDAC isoforms. For example the HDAC Class I specific inhibitor MS-275 is in advanced clinical trials (clinical trial Nos. NCT00020579, NCT00866333) for several forms of cancer, and the HDAC class II specific inhibitor ACY-1215 is at an advanced clinical phase (clinical trial Nos. NCT01323751, NCT01583283) for myeloma.\n\nHDAC inhibitors have been shown to synergize the actions of antifungal agents, due to their effect on preventing drug resistance in vitro17,18. Therefore, an alternative approach to address fungal drug resistance could be to harness the potential of modulating fungal gene expression by inhibition of fungal HDACs17,19.\n\nHos2, a class I HDAC enzyme plays an important role in gene activation in the yeast Saccharomyces cerevisiae by binding to open reading frames (ORFs) of active genes20. Hos2/Set3 histone deacetylase complex (Set3C) plays a key role in the conversion of white phase to virulent opaque phase in C. albicans21. Deletion of Hos2, the catalytic subunit of the Set3 complex produced a phenotype resembling inhibition of the Set3C by Trichostatin-A (TSA)22. Serum-induced morphogenesis of some C. albicans strains was shown to be inhibited by TSA19. Thus inhibiting the morphogenetic ability of this opportunistic pathogen using HDAC inhibitors holds the promise of future antifungal agents. Recently a small molecule, MGCD 290 (Hos2 inhibitor) has entered clinical trials (clinical trial number NCT01497223)) for use in combination with azoles, such as fluconazole, for fungal infections18.\n\nIn light of the emerging utility of Hos2 inhibition as an anti-fungal strategy, we have cloned and characterized the C. albicans HOS2. In this report, we present details on optimizing the codons and cloning for heterologous gene expression in Sf9 insect cells. Purified Hos2 was characterized by functional deacetylase activities on histone/tubulin preparations and inhibition studies with SAHA, TSA and MS-275. The Candida genome database reports Hos2 to be a putative NAD+ dependent histone deacetylase23, reminiscent of sirtuins. The Candida genome encodes for 3 sirtuins (HST1, HST2 and HST3). It has been shown that inhibition of HST3 by either gene repression or nicotinamide treatment reduces to a considerable extent the clinical severity of candidiasis24. In light of this reported observation, we checked for sirtuin like activity of purified Hos2 preparations using resveratrol, which activates sirtuins25.\n\n\nMaterials and methods\n\nC. albicans ATCC 90028 was obtained from ATCC and grown in Sabouraud dextrose media. The protoplasts were prepared from an overnight culture of fully-grown mycelia using zymolyase (Cat. No. L5263, Sigma, St. Louis, MO, USA) as per the manufacturer’s instructions. Protoplasts were lysed in a chaotrophic salt solution and genomic DNA was isolated according to manufacturer’s instruction (Qiagen, Hilden, Germany). The final eluate was reprecipitated with ammonium acetate and isopropanol and DNA quantified by UV spectrophotometer (Spectramax Gemini XS, Molecular devices, CA. USA).\n\nOligos were designed with codon changes made for 4th and 271st serine residues. Full length HOS2 gene was amplified by using 4 different primers (Table 1) using splicing by overlap extension (SOE PCR), so that codon usage could be maintained in any heterologous expression system. The full-length blunt end PCR product was cloned in to pJET1.2 cloning vector (CloneJET PCR Cloning Kit Thermo Scientific) and confirmed by restriction digestion using BamH1 and Not1. DNA sequencing using T7 promoter (forward, 5´-TAATAC GACTCACTATAGGG-3´) and pJET1.2 (reverse, 5´-AAGAACATCGATTTTCCATGGCAG-´3) sequencing primer ascertained the sequence of the recombinant Hos2 gene.\n\nRecombinant baculoviruses were generated using the Bac-to-Bac® baculovirus expression system according to the instructions of the manufacturer (Cat. No. 10359-016, Invitrogen, Carlsbad, USA). Sf9 insect cells (Cat. No. B825-01, Invitrogen, Carlsbad, USA) were cultured in complete TMN-FH medium (Cat. No. 554760, BD Bioscience, NJ, USA) served as host cells for virus generation and/or protein production. The HOS2 gene was cloned in frame with N-terminal hexa-histidine tag into the transfer vector pFastBac-HT B (Cat. No. 10584-027, Invitrogen, Carlsbad, USA) and transformed in to Escherichia coli DH10Bac cells (Cat. No. 10361-012, Invitrogen, Carlsbad, USA). Sf9 cells were transfected using cellfectin reagent (Cat. No. 10362-100, Invitrogen, Carlsbad, USA) with the recombinant bacmid, and the resulting viruses were tested for their ability to produce recombinant Hos2 protein using western blot. Production cultures were performed T-150 cell culture flasks (Greiner bio-one GmbH, Germany) at a density of 16×106 cells per flask. The cultures were inoculated with recombinant baculovirus stocks at multiplicities of infection of 10. At 4 days after infection, cells were harvested by centrifugation at 1200 g for 15 min at 4°C in a tabletop centrifuge (Model 5804R, Eppendorf AG, Hamburg, Germany). Cell pellets were stored at -80°C until protein purification.\n\nThe infected Sf9 cell pellets were resuspended in in-house ice-cold lysis buffer [10 mM Tris pH 7.5, 130 mM NaCl, 1% triton X-100, 10 mM NaF, 10 mM NaPi (sodium phosphate), and 10 mM NaPPi (sodium pyrophosphate)] with 1 X EDTA free protease inhibitor cocktail (Cat. No. 539134, Calbiochem, Merck-Millipore, San Diego, CA). Insoluble material was removed by centrifugation (Model 5804R, Eppendorf AG, Hamburg, Germany) at 16000 g, 4°C for 15 min. The cleared lysate was processed for nickel affinity chromatography. Briefly, the cell supernatant was loaded onto a 3-ml nickel-nitrilotriacetic acid-agarose resin (Ni-NTA agarose, Cat. No. 30210, Qiagen, Hilden, Germany) packed column pre-equilibrated with equilibration buffer (25 mM Tris-Cl pH 8.0, 300 mM NaCl). The column was washed with wash buffer (25 mM Tris-Cl pH 8.0, 300 mM NaCl, 20 mM imidazole). His-tagged Hos2 protein was eluted with buffer containing 300 mM imidazole (Sigma, St. Louis, MO, USA). The Hos2 protein preparation was dialyzed into buffer (25 mM Tris, pH 8.0, 138 mM NaCl, 10% glycerol) and kept in 100 μl aliquots at -70°C. The concentration of Hos2 protein in the final eluate was estimated by Bradford assay (Bio-Rad Laboratories, CA, USA).\n\nHDAC inhibitors, namely SAHA, TSA or MS-275 were dissolved in DMSO as 10 mM stock and subsequently diluted in 1X assay buffer (50 mM Tris Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 2.5 mM MgCl2, 1 mg/ml BSA). Enzymatic assay was carried out using the fluorogenic Class I HDAC substrate, Boc-Lys (Ac)-AMC (Cat. No. I1875, Bachem AG, Bubendorf, Switzerland). Briefly, 0.5 μg of purified recombinant protein in a volume of 10 μl of assay buffer was incubated with 50 μl of appropriate concentration of HDAC inhibitors and 20 μM of substrate at 37°C for 1 hr. Reactions were terminated by the addition of trichostatin A (TSA)/suberoylanilide hydroxamic acid (SAHA), trypsin (1 mg/mL) (Sigma, St. Louis, MO, USA) and left at 37°C for 15 min, before reading the plates in a fluorimeter (Spectramax Gemini XS, Molecular devices, CA. USA) at wavelengths 360 nm (ext) and 460 nm (emi).\n\nFemale BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, ME) at 4 weeks of age and then housed at the animal facility, Orchid chemicals and Pharmaceuticals for 2 weeks in a specific-pathogen free facility with a 12 h light cycle (6 am–6 pm) and a 12 h dark cycle (6 pm–6 am). Groups of four mice were housed in sterilised polypropylene cages covered with stainless steel grid top, lined with autoclaved clean rice husk bedding. All animal experimentations were approved by the institutional animal ethics committee (Protocol No. 01/IAEC-05/PPK/2009). The native Hos2 protein (expressed in pET-32 bacterial vector system and purified using nickel affinity chromatography under denaturating conditions) was emulsified in complete Freund’s adjuvant and injected subcutaneously into two female BALB/c mice (20 μg/mice). Booster doses of the deacetylase antigen (20 μg/mice) were given on the 14th and 21st day in Freund’s incomplete adjuvant. Mice were bled 7 days after the second booster dose and polyclonal anti sera was separated after clotting the blood26.\n\nJurkat, a human T lymphocyte cell line, and HeLa, a human cervical adenocarcinoma cell line was obtained from ATCC and were cultured in DMEM (Gibco, Life technologies) supplemented with 10% (v/v) fetal calf serum (FCS), 2 mM glutamine, 100 units/ml penicillin and 100 µg/ml streptomycin (Gibco, Life technologies).\n\nAcetylated histones were isolated from HeLa cells treated with the HDAC inhibitor SAHA as per published protocol27. The histone pellet was then resuspended in ultra pure water, stored in 50 μl aliquots at -70°C and the protein concentration was determined using a BCA kit (Pierce).\n\nC. albicans ATCC 90028 mycelia (~ 5 gm wet weight) were washed with water, centrifuged at 10,000 rpm for 10 minutes at 4°C and the mycelial pellet was resuspended in 50 ml of 0.1 mM Tris-HCl, pH 9.4, 10 mM DTT. The sample was incubated with shaking at 30°C for 15 min and pelleted. The pellet was washed with 50 ml of sorbitol/HEPES buffer (1.2 M sorbitol, 20 mM Hepes, pH 7.4) and left resuspended in the same buffer containing lyticase (1000 units) overnight at 30°C for spheroplasting28. The sample was pelleted and proceeded to histone isolation using acid extraction as described previously29. Acetone was added at 3:1 (vol/vol) to precipitate the histones, which were subsequently dissolved in 10 mM Tris, pH-8.0.\n\nAcetylated histones isolated from SAHA treated HeLa cells were used for the deacetylation assay with recombinant Hos2 enzyme. In brief, purified acetylated histones (2 μg) were incubated with different amounts of recombinant Hos2 enzyme. Histone deacetylation with 300 ng of rhHDAC1 or rhHDAC6 (expressed in-house using Baculo-viral expression system) was used as a positive control. Deacetylation assays were carried out for 1 hr at 37°C in reaction buffer (50 mM Tris Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 2.5 mM MgCl2, 1 mg/ml BSA). At the end of incubation, the reaction was stopped by the addition of 1X Lammeli sample buffer. The protein samples were resolved by SDS-PAGE and immunoblotted with anti-acetylated H3 Histone (Ac-K-9) antibody to study the deacetylation of H3-histone by rHos2.\n\nWhole cell extracts from TSA (Sigma, St. Louis, MO, USA) treated Jurkat cells were used for the α-tubulin deacetylation assay with recombinant Hos2 enzyme. In brief, whole cell extract (10 μg) were incubated with 5 and 8 μg of recombinant Hos2 protein. Deacetylation assays were carried out for 3 hr at 37°C in reaction buffer (50 mM Tris-Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 2.5 mM MgCl2). At the end of incubation, the reaction was stopped by the addition of 1X Lammeli sample buffer. The protein samples were resolved by SDS-PAGE and immunoblotted with either anti-acetylated α-tubulin or anti-α-tubulin antibodies.\n\nRecombinant Hos2 protein or the acetylated/deacetylated nuclear histones from deacetylation assays were separated on 10–15% SDS-polyacrylamide gels. The gels were stained with coomassie blue (Phast gel, Aersham) or electro-blotted onto nitrocellulose membrane (HyBond-C, Amersham), which was blocked with 5% nonfat dry milk (Cat No. 70166, Fluka, Sigma-Aldrich) in 0.1% Tween- 20 in TBS for 1 hr, followed by overnight incubation at 4°C with either the polyclonal mouse serum (dilution 1:4000 of the sera), rabbit polyclonal anti-acetylated H3-Histone (dilution 1:4000, Ac-K-9 of H3-Histone, Cat. No. 06-599; Millipore), mouse monoclonal anti-acetylated tubulin (dilution 1:15000, Cat. No. T7451, Sigma) or mouse monoclonal anti-tubulin antibody (dilution 1:150000, Cat. No. T6199, Sigma). After incubation with the appropriate horseradish peroxidase conjugated secondary antibodies (Bovine anti-mouse IgG HRP conjugate-dilution 1:7500, Santa Cruz, SC-2371, Goat anti-rabbit IgG-HRP Upstate (Millipore) 12–348, dilution 1:4000), supersignal west pico substrate (Thermo Scientific Pierce) was used for detection.\n\nHeLa nuclear extract (2 μg) or recombinant Hos2 enzyme (300 ng) was incubated with NAD+ (500 μM), Fluor de Lys®−Sirt1 (Enzo lifescience), varying concentrations of resveratrol (Cat. No. 0219605205, MP Biomedicals, Ohio, USA) (30, 100, 250 and 500 μM) in presence or absence of the pan-HDAC inhibitor trichostatin, at 37°C for 30 min. The reaction was carried out following manufacturer’s protocol (Enzo lifescience).\n\n\nResults\n\nCodon usage in C. albicans is different from standard genetic code. The 4th serine and the 271st serine of HOS2 are translated as leucine in both mammalian and in insect cell expression systems. Hence, these codons were mutated using oligonucleotide primers to express serine in the recombinant Hos2 protein. The PCR product was cloned in the pFastbac-HTB shuttle vector and subsequently into baculo viral DNA using the Bac-to-Bac expression system.\n\nThe Hos2 enzyme was expressed in the baculoviral-insect cell expression system as a NH2-terminal hexa histidine tagged fusion protein, which is detected on Western blot as a ~ 52 kDa protein using our own polyclonal anti-Hos2 anti-sera raised against Hos2 protein in mice. SDS-PAGE analysis of purified protein revealed a major band at ~ 52 kDa (Figure 1A).\n\n(A) Sf9 cells were infected with 10 MOI of recombinant Hos2 baculovirus for 72 hours and the soluble fraction was collected. Two μg of protein was separated on a 10% SDS-PAGE. Lane 2 is coomassie stained SDS-PAGE and Lane 3 is Hos2 protein recognized by polyclonal anti sera from mice. Molecular weight markers are labeled at the left side (Lane 1) of the blot. (B) Concentration dependent increase in deacetylase activity following incubation of baculo expressed Hos2 with Boc-lys (ac)-AMC fluorogenic peptide substrate in 1 hr assay at 37°C.\n\n\n\nRecombinant Hos2 enzyme was assayed for deacetylase activity using the synthetic deacetylase substrate, Boc-Lys (ac)-AMC. The total activity with Boc-Lys (ac) AMC showed the enzyme to be active in deacetylating the lysine residue and the activity increased with an increase in Hos2 concentration (Figure 1B).\n\nThe inhibition of deacetylation activity of recombinant Hos2 was studied using classical HDAC inhibitors namely SAHA, TSA and MS-275. TSA was very potent in inhibiting Hos2 with an IC50 of 2.8 nM, SAHA inhibited Hos2 with an IC50 of 65 nM (Figure 2, Table 2). However, MS-275 showed > 50% inhibition of Hos2 activity only at 10 μM (Table 3).\n\nRecombinant Hos2 enzyme was incubated with different concentrations of pan-HDAC inhibitors (SAHA or TSA) and the enzyme activity assay was performed.\n\nThe ability of purified Hos2 protein to deacetylate acetylated histones was examined in vitro using acetylated nuclear histone preparation made from SAHA treated HeLa cells. The nuclear histones from HeLa cells were isolated using a modified protocol of Shechter et al.27 and established the deacetylation assay using rhHDAC1/rhHDAC6 as controls along with rHos2. In these assays it was found that 0.3 μg of rhHDAC1 was able to deacetylate the nuclear acetylated histones as detected by an anti-H3-K9 histone antibody. Recombinant hHDAC1 was more potent in deacetylating lysine residues in H3-histones than rhHDAC6. However, no significant deacetylation of H3-Histone was seen with 0.3 μg of Hos2 (Figure 3A). Similar results were obtained with acetylated histones from Candida in the in vitro histone deacetylation assay with purified Hos2 protein (up to 3 μg, Figure 3B). In contrast, when recombinant Hos2 was incubated with mammalian acetylated α-tubulin, a significant reduction in acetylation of α-tubulin (K40) could be observed (Figure 4).\n\nA. Histone H3 acetylation determined by Western blot of extracted acetylated histones (2 µg) Control (Lane 1), incubated with 300 ng of recombinant Hos2 protein (Lane 2), recombinant human HDAC1 (HD1 Lane 3) and recombinant human HDAC6 (HD6 Lane 4) for 1 hour at 37°C in HDAC assay buffer. B. Histone H3 acetylation determined by Western blot of extracted acetylated histones (2 µg) isolated from C. albicans, incubated with different concentrations of recombinant Hos2 protein for 1 hour at 37°C in HDAC assay buffer.\n\nα-tubulin acetylation determined by Western blot of SAHA-treated whole cell extracts containing acetylated α-tubulin (10 µg) incubated with 0, 5 or 8 µg of recombinant Hos2 for 3 hrs at 37°C in HDAC assay buffer.\n\n\n\nSince Hos2 is a putative NAD+ dependent deacetylase, a feature of sirtuin class of deacetylases, it was of interest to check if the Hos2 protein displayed any sirtuin like activity. The sirtuin-like activity in response to the sirtuin activator resveratrol was studied using fluor-de-lys, a synthetic Sirt1 substrate. HeLa nuclear extract was used as a positive control for sirtuin activity. In the presence of TSA where the HDAC activities are inhibited no appreciable NAD+ dependent deacetylase-like activity was seen, following incubation of Hos2 with different concentrations of resveratrol (Figure 5).\n\nFold activation of the Sirt1 like activity following incubation of recombinant Hos2 enzyme with different concentration of resveratrol in presence and absence of HDAC inhibitor trichostatin.\n\n\n\n\nDiscussion\n\nHos2 is a histone deacetylase and interacts with several proteins both in the cytoplasmic milieu as well in the nucleus30. Hos2 along with Set3 (SET domain-containing protein 3) an associated protein, plays an important role in controlling gene expression by associating with transcriptionally active regions of the chromatin31.\n\nHDAC inhibitors by virtue of their ability to prevent antifungal resistance in vitro have been proposed as antifungal adjuvants. It has been surmised that inhibiting Hda1 for example might enhance the anti-fungal effect of HDAC inhibitors by limiting hyphal development, while inhibiting Hos2 might contribute to limiting yeast development32. Hos2 has an essential function in morphogenesis especially during conditions of nitrogen starvation33. The critical role of HDAC`s in C. albicans pathogenesis and survival to antifungal treatment underscores the necessity to study HDAC function in this organism.\n\nThe increasing clinical incidences of azole resistant fungal infections in critical care patients, makes a good reason to find additional drug targets to control such diseases. There is at least one small molecule (MGCD 290) that inhibits Hos2 histone deacetylase that has progressed to clinical trials. MGCD 290 in combination with azoles was shown to be active against azole resistant yeasts and moulds18.\n\nIn order to better understand the role played by the Candida Hos2 enzyme we attempted to clone, express and characterize the protein in detail. This study describes the cloning, expression, purification and characterization of the Hos2 deacetylase enzyme from C. albicans.\n\nWe cloned and expressed the HOS2 gene in baculoviral expression system as a 6x his-tagged protein, which exhibits classical deacetylase activity with the synthetic Boc-Lys (ac)-AMC peptide substrate. In our study, the yield of the Hos2 protein was generally low and probably could be attributed to difference in codon usage between Candida and Sf9 insect cells. This in vitro enzymatic assay, amenable to high throughput, could be used for screening potent fungal Hos2 inhibitors that could be a potential anti-fungal adjuvant.\n\nOur studies with the recombinant Hos2 protein showed that it is susceptible to inhibition by standard HDAC inhibitors such as SAHA and TSA. We characterized the inhibition profile of purified proteins with SAHA and TSA and showed that TSA is a more potent inhibitor of Hos2 with an IC50 of 2.8 ± 0.9 nM compared to SAHA (IC50 65.4 ± 2.4 nM). Our studies with the class I HDAC inhibitor MS-275 showed that this inhibitor did not inhibit Hos2 deacetylase as effectively as the pan HDAC inhibitors SAHA or TSA, suggesting that Candida Hos2 is more similar to class II deacetylases.\n\nThe recombinant Hos2 failed to deacetylate either mammalian or fungal nuclear histones, suggesting that the histones are not the preferred substrates for the Hos2 enzyme. Experiments with total lysates from Jurkat cells containing acetylated α-tubulin showed a dose dependent deacetylation albeit at higher concentration of Hos2 (> 5 μg). Hos2 in essence resembles the class II mammalian HDACs, specifically HDAC6 in its preference for tubulin deacetylation. The physiological relevance of tubulin deacetylation by Hos2 warrants further study.\n\nThe Candida genome database23 predicts Hos2 protein to be a NAD+ dependent deacetylase. Sirtuins which are classified as Class III deacetylases are NAD+ dependent enzymes activated by polyphenols such as resveratrol. Sirtuins have been proposed to regulate cellular metabolism, ageing and other related processes, specifically cellular stress response to caloric restriction, mediating life span extension. The role of resveratrol as a sirtuin activator has been resolved recently and it is now known that in addition to activating Sirt1, it also activates Sirt5, while inhibiting Sirt325. Thus inhibiting any sirtuin like activity with small molecule inhibitors could be another way of enhancing the activity of currently used anti-fungals. We evaluated the possibility of Hos2 being a sirtuin like enzyme with a known Sirt1 activator resveratrol. We did not observe any significant activation of NAD+ dependent deacetylase activity with the fluor-de-lys substrate.\n\nIn conclusion this study establishes a functional assay for purified Hos2 protein. This in vitro enzymatic assay can be used to screen small molecule inhibitors of Hos2, which can synergise current anti-fungals in the clinic.", "appendix": "Author contributions\n\n\n\nGK, MPS and SN conceived the study. GK, NDM and RG designed the experiments. GK, NDM and RG carried out the research. GK and NDM prepared the first draft 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 was an exploratory in-house project funded solely by Orchid Chemicals and Pharmaceuticals Limited.\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. B. Gopalan, Chief Scientific Officer (CSO) for the encouragement and support and the Management of Orchid Chemicals and Pharmaceuticals Limited (www.orchidpharma.com), Chennai, India, for encouraging publication of this research.\n\n\nReferences\n\nOdds FC: Candida and candidosis: a review and bibliography: Bailliere Tindall; 1988. Reference Source\n\nPfaller MA, Diekema DJ, Jones RN, et al.: International surveillance of bloodstream infections due to Candida species: frequency of occurrence and In vitro susceptibilities to fluconazole, ravuconazole, and voriconazole of isolates collected from 1997 through 1999 in the SENTRY antimicrobial surveillance program. J Clin Microbiol. 2001; 39(9): 3254–3259. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFera MT, La Camera E, De Sarro A: New triazoles and echinocandins: mode of action, In vitro activity and mechanisms of resistance. Expert Rev Anti Infect Ther. 2009; 7(8): 981–998. PubMed Abstract \| Publisher Full Text\n\nSanglard D, Ischer F, Koymans L, et al.: Amino acid substitutions in the cytochrome P-450 lanosterol 14alpha-demethylase (CYP51A1) from azole-resistant Candida albicans clinical isolates contribute to resistance to azole antifungal agents. Antimicrob Agents Chemother. 1998; 42(2): 241–253. PubMed Abstract \| Free Full Text\n\nWhite TC: The presence of an R467K amino acid substitution and loss of allelic variation correlate with an azole-resistant lanosterol 14alpha demethylase in Candida albicans. Antimicrob Agents Chemother. 1997; 41(7): 1488–1494. PubMed Abstract \| Free Full Text\n\nSong JL, Harry JB, Eastman RT, et al.: The Candida albicans lanosterol 14-alpha-demethylase (ERG11) gene promoter is maximally induced after prolonged growth with antifungal drugs. Antimicrob Agents Chemother. 2004; 48(4): 1136–1144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSanglard D, Kuchler K, Ischer F, et al.: Mechanisms of resistance to azole antifungal agents in Candida albicans isolates from AIDS patients involve specific multidrug transporters. Antimicrob Agents Chemother. 1995; 39(11): 2378–2386. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlbertson GD, Niimi M, Cannon RD, et al.: Multiple efflux mechanisms are involved in Candida albicans fluconazole resistance. Antimicrob Agents Chemother. 1996; 40(12): 2835–2841. PubMed Abstract \| Free Full Text\n\nMarr KA, Lyons CN, Rustad TR, et al.: Rapid, transient fluconazole resistance in Candida albicans is associated with increased mRNA levels of CDR. Antimicrob Agents Chemother. 1998; 42(10): 2584–2589. PubMed Abstract \| Free Full Text\n\nClayton AL, Hazzalin CA, Mahadevan LC: Enhanced histone acetylation and transcription: a dynamic perspective. Mol Cell. 2006; 23(3): 289–296. PubMed Abstract \| Publisher Full Text\n\nMarks PA: Discovery and development of SAHA as an anticancer agent. Oncogene. 2007; 26(9): 1351–1356. PubMed Abstract \| Publisher Full Text\n\nCampas-Moya C: Romidepsin for the treatment of cutaneous T-cell lymphoma. Drugs Today (Barc.). 2009; 45(11): 787–795. PubMed Abstract \| Publisher Full Text\n\nDuvic M, Vu J: Update on the treatment of cutaneous T-cell lymphoma (CTCL): Focus on vorinostat. Biologics. 2007; 1(4): 377–392. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDokmanovic M, Clarke C, Marks PA: Histone deacetylase inhibitors: overview and perspectives. Mol Cancer Res. 2007; 5(10): 981–989. PubMed Abstract \| Publisher Full Text\n\nMai A: The therapeutic uses of chromatin-modifying agents. Expert Opin Ther Targets. 2007; 11(6): 835–851. PubMed Abstract \| Publisher Full Text\n\nRiester D, Hildmann C, Schwienhorst A: Histone deacetylase inhibitors--turning epigenic mechanisms of gene regulation into tools of therapeutic intervention in malignant and other diseases. Appl Microbiol Biotechnol. 2007; 75(3): 499–514. PubMed Abstract \| Publisher Full Text\n\nSmith WL, Edlind TD: Histone deacetylase inhibitors enhance Candida albicans sensitivity to azoles and related antifungals: correlation with reduction in CDR and ERG upregulation. Antimicrob Agents Chemother. 2002; 46(11): 3532–3539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPfaller MA, Messer SA, Georgopapadakou N, et al.: Activity of MGCD290, a Hos2 histone deacetylase inhibitor, in combination with azole antifungals against opportunistic fungal pathogens. J Clin Microbiol. 2009; 47(12): 3797–3804. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimonetti G, Passariello C, Rotili D, et al.: Histone deacetylase inhibitors may reduce pathogenicity and virulence in Candida albicans. FEMS Yeast Res. 2007; 7(8): 1371–1380. PubMed Abstract \| Publisher Full Text\n\nWang A, Kurdistani SK, Grunstein M: Requirement of Hos2 histone deacetylase for gene activity in yeast. Science. 2002; 298(5597): 1412–1414. PubMed Abstract \| Publisher Full Text\n\nHnisz D, Schwarzmuller T, Kuchler K: Transcriptional loops meet chromatin: a dual-layer network controls white-opaque switching in Candida albicans. Mol Microbiol. 2009; 74(1): 1–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHnisz D, Majer O, Frohner IE, et al.: The Set3/Hos2 histone deacetylase complex attenuates cAMP/PKA signaling to regulate morphogenesis and virulence of Candida albicans. PLoS Pathog. 2010; 6(5): e1000889. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArnaud MB ID, Skrzypek MS, Binkley J, et al.: Candida Genome Database. Reference Source\n\nWurtele H, Tsao S, Lepine G, et al.: Modulation of histone H3 lysine 56 acetylation as an antifungal therapeutic strategy. Nat Med. 2010; 16(7): 774–780. PubMed Abstract \| Publisher Full Text\n\nGertz M, Nguyen GT, Fischer F, et al.: A molecular mechanism for direct sirtuin activation by resveratrol. PLoS One. 2012; 7(11): e49761. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFuller SA, Takahashi M, Hurrell JG: Immunization of Mice. Curr Protoc Mol Biol. John Wiley & Sons, Inc.; 2001. PubMed Abstract \| Publisher Full Text\n\nShechter D, Dormann HL, Allis CD, et al.: Extraction, purification and analysis of histones. Nat Protoc. 2007; 2(6): 1445–1457. PubMed Abstract \| Publisher Full Text\n\nYeast Nuclei Isolation. Reference Source\n\nKnapp AR, Ren C, Su X, et al.: Quantitative profiling of histone post-translational modifications by stable isotope labeling. Methods. 2007; 41(3): 312–319. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWiren M, Silverstein RA, Sinha I, et al.: Genomewide analysis of nucleosome density histone acetylation and HDAC function in fission yeast. EMBO J. 2005; 24(16): 2906–2918. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHnisz D, Bardet AF, Nobile CJ, et al.: A histone deacetylase adjusts transcription kinetics at coding sequences during Candida albicans morphogenesis. PLoS Genet. 2012; 8(12): e1003118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZacchi LF, Schulz WL, Davis DA: HOS2 and HDA1 encode histone deacetylases with opposing roles in Candida albicans morphogenesis. PLoS One. 2010; 5(8): e12171. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobyr D, Suka Y, Xenarios I, et al.: Microarray deacetylation maps determine genome-wide functions for yeast histone deacetylases. Cell. 2002; 109(4): 437–446. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4016", "date": "19 Mar 2014", "name": "Donna MacCallum", "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 an interesting study designed to examine the biological activities of Candida albicans Hos2 enzyme, a putative histone deacetylase. The authors express a recombinant version of the protein, which has a His tag added to the N-terminus, and use this to explore potential substrates for the enzyme.There are a number of major issues with the data as presented, which need to be addressed:It is difficult to make any conclusions from the quantitative data presented in the figures as there are no error bars and the number of replicates is not stated. In addition, it is unclear what the data in Figures 1B and 5 has been expressed relative to (i.e. fold change relative to...?) Statistical analyses should also be carried out to determine whether any differences are statistically significant ones. The authors should also revisit the data presented in Figure 2 and Table 2. The data does not correlate, as the authors suggest that the IC50 values are in micromolar amounts, yet the data in the figure would suggest millimolar levels - units should be checked.The title for Figure 4 should be modified, since this figure contains only data for beta-tubulin. In the main text the authors need to be much clearer regarding the rationale for looking at tubulin as a substrate. The authors may also wish to consider the possibility that the substrate specificity and/or activity may have been affected by the N-terminal tag and how this could be investigated. The discussion section should be revisited to remove the reiteration of the introduction and results. Instead, use this part of the paper to discuss the findings in relation to other published work. Minor points:Candida albicans is capable of causing superficial (oral and vaginal thrush) infections in immune normal hosts (abstract and introduction).Gene names, e.g. CDR1 and ERG11, should be in italics.For the enzymatic assay it would be good to state the volume of the assay so that final concentrations can be worked out.Figure and table legends should include the number of replicates carried out to generate the data.", "responses": [ { "c_id": "821", "date": "13 May 2014", "name": "Karthikeyan Ganesan", "role": "Author Response", "response": "Error bars have been incorporated and where possible, statistical analysis done and reported.Error in reporting IC50 value in Table 2 and Fig. 2 is corrected.Title for Fig. 4 changed to reflect tubulin deacetylation.The rationale for choosing tubulin as a substrate has been explained. We agree with the reviewer that N-terminal tag could perhaps play a role, although we consider that as a remote possibility.Minor points suggested by reviewer has also been addressed." } ] }, { "id": "4017", "date": "16 Apr 2014", "name": "David Soll", "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 an in vitro assay strategy for testing the effects of antifungal compounds on the activity of Hos2 deacetylase large scale screens.The goal is to find compounds that could be used as facilitators in combinatorial drug treatments for azole resistant Candida albicans infections. Although in vitro assays are sometimes useful for developing in vivo strategies, they are not ideal for screening and identifying optimal drugs.As shown in this study by the authors, the recombinant Hos2 they generated exhibits paradoxically tubulin-specific deacetylase activity and not Hos2 specific activity. This finding seems to undermine their intent, because anti-tubulin deacetylase activity or anti-human HDAC6 activity demonstrated in vitro could be an artifact of the recombinant protein. In addition, anti-tubulin and anti-HDAC6 activity might result in adverse side effects on human host functions and thus affect the specificity of the primary drug.It is surprising that the recombinant Hos2 did not show any activity in acetylating either the human or Candida histone H3 in vitro, despite the fact that several earlier studies in C. albicans has demonstrated that Hos2 is involved in morphogenetic programs by affecting chromatin function. In addition, an isolated study has demonstrated the efficacy of a Hos2-specfic inhibitor on azole-resistant strains of C. albicans (Pfaller et al., 2009), which is quoted by the authors. The authors do state in the introduction that Hos2 is not a typical deacetylase and has been described by the Candida Genome Database as a member of NAD-dependent sirtuins. The authors’ data show that their recombinant version lacks this activity, but does exhibit inhibition by classical inhibitors of histone deacetylases.In addition to this major problem there are a few other aspects of the paper that need to be addressed:Firstly, Figure 1 and Figure 5 do not have error bars to indicate reproducibility of the data. Also, a statistical analysis of data sets would be useful to indicate the significance of the results. Secondly, the units in Figure 2 and Table 2 need to be consistent with the units described in the text. Thirdly, since Figure 4 shows the effect of recombinant Hos2 on only tubulin, the title needs to be changed to indicate that the assay was not done with purified histones. Finally, it is not clear what “fold activity” (in Figure 1B) and “fold activation” (Figure 5) refer to.", "responses": [ { "c_id": "819", "date": "13 May 2014", "name": "Karthikeyan Ganesan", "role": "Author Response", "response": "As a pharmaceutical organization, involved in drug discovery, our initial compound libraries are screened for in vitro activity/inhibition assays. These in vitro assays act as important filters removing in-active compounds to progress only the most potent ones to in vivo screening.Deacetylase assays using recombinant Hos2 enzyme and synthetic substrate Boc (Lys) AMC showed that the enzyme is active in standard fluorogenic assays. Absence of deacetylase activity with histone preparations (Human/Candida) in the in vitro histone deacetylation assays was a surprise but the fact that Class I HDAC inhibitor like MS-275 did not inhibit rHos2 in standard fluorogenic assays, led us to explore alternate substrate, namely tubulins.We do agree with the reviewer that anti-tubulin deacetylase activity could be an artifact of recombinant protein, but we strongly feel that it is a remote possibility.Any small molecule that shows anti-tubulin or anti-HDAC6 activity would be picked up in the in vitro screen and would be prevented from moving further up the discovery path. (Which can be re-purposed for different therapeutic )We are indeed aware of the previous studies showing the effect of Set3/Hos2 complex and the opposing roles of Hda1 and Hos2 in Candida morphogenesis as well as the role of Hos2 in S. cerevisiae. In these studies Histone H4 acetylation levels vary dramatically, especially H4-K16, however the Histone H3 levels more or less remain constant, (supplementary materials from Wang et al., 2002) and that`s probably the reason why we don`t see any dramatic deacetylation when probed with Histone H3-K9. The study by Pfaller et al. did have in vivo data, but did not have any in vitro data.Error bars and statistical analysis have been incorporated for Figs. 1 and 5. Units for Fig. 2 and Table 2 corrected.Label has been changed for Fig. 4Fold activity and Fold activation is with respect to Blank and has been reflected in the figures." } ] } ]	1	https://f1000research.com/articles/2-238
https://f1000research.com/articles/3-165/v1	22 Jul 14	{ "type": "Research Article", "title": "Effect of anesthesia level on murine cardiac function", "authors": [ "Fabien Picard", "Francois Depret", "Sergio Zanotti-Cavazzoni", "Steven Hollenberg", "Sergio Zanotti-Cavazzoni", "Steven Hollenberg" ], "abstract": "Background: Echocardiography allows for sensitive and non-invasive assessment of cardiac function in mice, but requires sedation and immobility, which influences cardiac performance. Minimizing the hemodynamic effects of anesthesia is extremely important for improving the applicability of animal models to the clinical setting. We sought to evaluate the effects of isoflurane dose on myocardial function in a murine model.Methods: Twelve healthy C57BL/6 mice were studied with three different isoflurane anesthesia regimens: deep anesthesia with an objective of heart rate (HR) between 350 and 400 beats per minute (bpm), light anesthesia with an objective of HR between 475 and 525 bpm and just before the mice woke up (>575 bpm). Using a high-resolution ultrasound system, stroke volume, cardiac output, left ventricle dimension and fractional shortening were recorded.Results: Fractional shortening was not statistically different in the awake group and the light anesthesia group (49±5% in awake mice vs. 48±5%; p=0.62), whereas it was different compared to the deep anesthesia group (31±5%, p<0.0001 compared to both groups). Similar results were found for stroke volume (41.4±5.8 ml vs. 41.6±6.9 ml; p=0.81 and 35±8.3 ml; p<0.05 compared to both groups). Cardiac output was slightly lower in the light anesthesia group compared to the awake group (21.9±3.6 ml/min vs. 25.6±3.3; p=0.02) due to HR significant difference (522±17 bpm vs. 608±23 bpm; p<0.0001).Conclusions: Doppler echocardiography can be performed under very light anesthesia using small doses of isoflurane without influencing cardiac inotropic function. This technique allows for accurate and reproducible assessment of cardiac function while minimizing hemodynamic perturbations.", "keywords": [ "Animal models of disease can provide important insights into pathophysiological mechanisms and allow for evaluation of novel therapies. Murine models have been widely used for these purposes: cardiac and vascular morphologic differences between mice", "rats", "and humans are fairly subtle", "and many physiological characteristics are comparable1." ], "content": "Introduction\n\nAnimal models of disease can provide important insights into pathophysiological mechanisms and allow for evaluation of novel therapies. Murine models have been widely used for these purposes: cardiac and vascular morphologic differences between mice, rats, and humans are fairly subtle, and many physiological characteristics are comparable1.\n\nFor hemodynamic assessments in murine models, it is critical to develop approaches for accurate and reproducible measurements of cardiac morphology and function in intact animals. The parameters measured in mice are close to those predicted by allometric formulas when compared to other mammals, supporting the view that physiological mechanisms in mice are closely related to those in humans. Echocardiography (echo) is convenient, quick, and safe and allows for consecutive and repeated evaluations of cardiovascular physiologic and pathologic characteristics in live animals. Echo has been widely applied in determining cardiac phenotypes and functions in murine models2–7. Anesthesia is necessary to achieve the best technical results but has a potential downside in that it may induce myocardial depression. Minimizing the hemodynamic effects of anesthesia is extremely important for improving the applicability of animal models to the clinical setting, especially in models of shock in which hemodynamic perturbations are paramount.\n\nOne of the classical models of septic shock in mice is the cecal ligation and puncture model8; the anesthetic usually used in this model is ketamine. Ketamine is well known to produce profound bradycardia with effects on loading conditions and ventricular function, and to be more difficult to control once injected9. Compared to ketamine, inhalation anesthesia with isoflurane has currently been considered ideal for experimental studies in the mouse because of its rapid induction, easy control of the depth of anesthesia, and relatively stable heart rates (HR) and blood pressure during observations. It also seems to have the least myocardial depression compared to other anesthesia regimens10–13. Few data concerning echocardiographic evaluation of left ventricle (LV) function in anesthetized and non-anesthetized mice are available7. It is also known that deeper levels of anesthesia lower heart rate and produce more myocardial depression14. In fact heart rate can be used as a proxy for depth of anesthesia.\n\nSome investigators have trained mice in order to perform echocardiography on awake mice, thus avoiding the potentially confounding effects of anesthesia7. The mice, however, need to be restrained, and these experiments induce adrenergic stress that may produce hemodynamic perturbations in their own right.\n\nThe optimal depth of anesthesia for reliable and relevant measurements by high-resolution echocardiography is not well studied. The aim of this study was to evaluate the effects of isoflurane dose on myocardial function in a murine model and to compare very low doses of isoflurane to a state in which the mice were off anesthesia. To evaluate whether a very light anesthesia using a small dose of isoflurane affected systolic cardiac function, we recorded cardiac performance in mice by using a high-resolution ultrasound system in three groups with different anesthesia regimens. The three groups were based on the depth of anesthesia corresponding to HR measurement: a deep anesthesia group corresponding to a low HR (350–400 bpm), light anesthesia corresponding to a high HR (475–525 bpm) and an awake group in which cardiac performance was assessed just before the animals woke up (HR>575 bpm). Left ventricle (LV) dimensions, systolic function and aortic pulsed wave (PW) Doppler were recorded.\n\n\nMaterials and methods\n\nA total of 12 healthy male C57BL/6J mice (10–12 weeks old, Jackson Laboratories, Bar Harbor, ME) weighing 26 to 30 g were included in this study. They were housed in the Central Animal Facility of the University of Medicine and Dentistry of New Jersey (UMDNJ), Camden, NJ, USA, at 20°C at 60% humidity with a 12:12-hr light–dark cycle and fed on a standard diet and water ad libitum, for at least seven days before experiments, to avoid preconditioning. Animal experiments were performed in accordance with the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (revised 1996; National Institutes of Health, Bethesda, MD), and the Animal Care and Use Committee of UMDNJ approved the study protocol. Animals were returned to the vivarium for other experiments after the study.\n\nTo assess the effect of different anesthesia regimens on cardiac function using isoflurane, we performed 36 echocardiographic studies on 12 mice and used HR as a proxy for the depth of anesthesia.\n\nWe first performed echocardiography on 12 mice with a deep anesthesia regimen corresponding to a low HR (350–400 bpm), and did a second echocardiographic study on the same mice three days later under light anesthesia titrated to a higher HR (475–525 bpm). The anesthesia was then discontinued and the same parameters were recorded continuously until the animals woke up, when they generally moved and the images were lost. The images taken just prior to movement were used for the awake group. The HRs in the awake group were regularly higher than HR>575 bpm. The ranges of high and low HRs were determined according to previous reports on the relationship between the HR and cardiac function15.\n\nIsoflurane induction was performed in an induction box with 3% isoflurane (Baxter) in pure medical oxygen. After the righting reflex disappeared, the animal was fixed in supine position on a heating pad (Vevo® Integrated Rail System, Visualsonics, Inc) to maintain normothermia and electrocardiographic limb electrodes were placed. The mouse was allowed to breathe spontaneously, and the chest was shaved to minimize ultrasound attenuation. Acoustic coupling gel (Aquasonic® 100, Parker Laboratories, Inc) was applied to the thorax surface to optimize the visibility of the cardiac chambers and wall movements. Anesthesia was maintained with 2% isoflurane for the low HR group and with 0.5% isoflurane for the high HR group. If the HR was found to be above the required range, the isoflurane concentration could be temporarily increased to 4% and then decreased to 2% after the HR reached the required level. Contrarily, if the HR was below the required range, the isoflurane concentration could be set at 0.25% and then increased.\n\nA Vevo® 770 high-resolution ultrasound system (VisualSonics, Inc) equipped with a 30-MHz, 100-frame-per-second micro-visualization scan head was used to perform echocardiography. The echocardiographic measurements were recorded according to standard methods from previously published reports7,11,16,17. A parasternal long-axis B-mode image was acquired with appropriate positioning of the scan head so that the maximum LV length could be identified, then a clockwise 90° rotation at the papillary muscle level was performed to obtain the parasternal short-axis view. The M-mode cursor was positioned perpendicular to the anterior and posterior walls of the LV. From this view, wall thickness and chamber dimensions were measured. Image loops were captured and included at least ten cardiac cycles. Data were averaged from at least two cycles per loop. End-diastole or end-systole was defined as the maximal or minimal LV diastolic or systolic diameter, respectively. The parameters obtained from M-mode tracings included the LV anterior wall end-diastolic thickness (LVAWTd), LV anterior wall end-systolic thickness (LVAWTs), LV posterior wall end-diastolic thickness (LVPWTd), LV posterior wall end-systolic thickness (LVPWTs), LV end-diastolic diameter (LVEDD) and LV end-systolic diameter (LVESD). Other parameters such as the LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), ejection fraction (EF), fractional shortening (FS) were calculated using the following formulas (VisualSonics Vevo® 770 Imaging System, Operator Manual):\n\n(1) LVEDV = [7.0 ÷ (2.4 + LVEDD)] × LVEDD3;\n\n(2) LVESV = [7.0 ÷ (2.4 + LVESD)] × LVESD3;\n\n(3) %EF = 100 × [(LVEDV − LVESV) ÷ LVEDV];\n\n(4) %FS = 100 × [(LVEDD − LVESD) ÷ LVEDD];\n\nAortic flow velocity time integral (AV VTI) images were measured with PW Doppler imaging in the suprasternal orientation. The Doppler angle was modulated slightly to allow the interrogation beam almost parallel to aortic flow so that the maximum flow velocity was obtained. PW Doppler signals from at least three cardiac cycles were averaged. Stroke volume (SV) and cardiac output (CO) were calculated using the following formulas (VisualSonics Vevo® 770 Imaging System, Operator Manual):\n\n(5) SV = [(aortic root diameter)2 × velocity-time integral × π/4];\n\n(6) CO = SV × HR;\n\nTo determine the intra-observer variability of echocardiographic measurements, one observer repeated the echocardiographic data analysis for eight randomly chosen mice on different days. Two observers measuring the already recorded image estimated inter-observer variability. Intra-observer and inter-observer errors were calculated as the difference between the two observations divided by the mean and expressed as a percentage. Intra- and inter-class correlation coefficients (ICC) were used to evaluate the agreement for both intra- and inter-observer variability. The ICC value is the ratio of the between-subject variance to the sum of the between-subject variance and the within-subject variance. The ICC values can be considered as poor (less than 0.40), fair (0.40–0.59), good (0.60–0.74), and excellent (0.75–1.00)18.\n\nData were averaged and reported as mean ± SD unless otherwise stated. Statistical analysis was performed using SPSS software (Version 19, SPSS, Chicago). Whenever appropriate, the data for the awake and the anesthetized animals were compared with Student’s t-test for matched pairs. Two-sided p-values <0.05 were considered as significant. Paired t-tests were used for intra-observer and inter-observer comparisons. Variation was evaluated by determination of ICC. This measure indicates excellent agreement if >0.75, fair to good if between 0.4 and 0.75, and poor if <0.4.\n\n\nResults\n\nThere were no differences in the baseline HR among the mice we tested.\n\nIn the deep anesthesia group, HR was significantly lower than in the light and awake group (363±27 bpm vs. 522±17 and 608±23 bpm; p<0.0001, Table 1, Figure 1). In addition, there was a significant reduction in FS between the deep anesthesia group and both the light and the awake group (31±5% vs. 48±5 and 49±5%; p<0.0001). EF was also lower in the deep anesthesia group, as compared to the light and to the awake group (59±7% in the deep anesthesia group and 80±5% and 81±5% in the light group and the awake group respectively; p<0.0001), as AV VTI (2.74+/-0.59 cm in the deep anesthesia group and 3.31±0.45 cm in the awake group; p=0.048), SV (35.01±8.27 μl in the deep anesthesia group and 41.35±5.79 μl in the awake group; p=0.048) and CO (12.67±3.08 ml/min. in the deep anesthesia group and 25.61±3.31 ml/min. in the awake group; p<0.0001) (Table 1, Figure 1).\n\nHR= Heart rate; LVAWTd/s=Left ventricle anterior wall end-diastolic/systolic thickness; LVEDD= Left ventricle end-diastolic diameter; LVPWTd/s= Left ventricle posterior wall end-diastolic/systolic thickness; LVESD= Left ventricle end-systolic diameter; LVEDV= Left ventricle end-diastolic volume; LVESV= Left ventricle end-systolic volume; EF= ejection fraction; FS= fractional shortening; CO= cardiac output;. Values are means +/- SD. N = 12 animals for the three different anesthetic regimens. #p < 0.05 compared to deep anesthesia. ‡p < 0.05 compared to light anesthesia. p < 0.05 compared to awake mice.\n\nCardiac performance in animals anesthetized with isoflurane in a deep anesthesia, a light anesthesia or just before waking. Heart rate (A), fractional shortening (B), cardiac output (C), stroke volume (D) were studied.\n\nN = 12 animals for the three different anesthetic regimens.\n\n#p < 0.05 compared to deep anesthesia. ‡p < 0.05 compared to light anesthesia.\n\np < 0.05 compared to awake mice.\n\nLVEDD was slightly higher on the deep anesthesia group (3.74±0.39 mm) but no statistically significant difference could be found between the light anesthesia group (3.0±0.33 mm; p=0.0827) and the awake group (3.37±0.41, p=0.0864) (Table 1).\n\nHR was significantly lower in the light anesthesia group compared to the awake group (522±17 bpm vs. 608±23 beats/min; p<0.0001) (Table 1, Figure 1). Concerning the FS values, there were no statistical differences between the light anesthesia group and the awake group (48±5% vs. 49±5%; p=0.6212). The same results were found for EF (80±5% in the light anesthesia group; p=0.5536 when compared to the awake group), AV VTI (3.26+/-0.5 cm in the light anesthesia group; p=0.8 when compared to the awake group) and SV (41.61+/-6.86 μl in the light anesthesia group; p=0.81 when compared to the awake group).\n\nAlthough SV was similar in both groups, CO was significantly lower in the light anesthesia group, (21.90+/-3.56 ml/min. in the light anesthesia group; p=0.02 when compared to the awake group) probably due to the significant difference in HR (Table 1, Figure 1).\n\nThe reproducibility of echocardiographic measurements [LVAWTd (mm), LVESD (mm), LVPWTd (mm), LVAWTs (mm), LVEDD (mm), LVPWTs (mm), FS (%), Peak AV (mm/s), AV VTI (cm)] in our study was good to excellent (Table 2) for both intra-observer variability (correlation coefficient: 0.95–0.99) and inter-observer variability (correlation coefficient: 0.67–0.99).\n\nLVEDD= Left ventricle end-diastolic diameter; LVESD= Left ventricle end-systolic diameter; LVAWTd/s=Left ventricle anterior wall end-diastolic/systolic thickness; LVPWTd/s= Left ventricle posterior wall end-diastolic/systolic thickness; FS= fractional shortening; AV VTI= Aortic valve velocity time integral.\n\nValues are means +/-SEM; ICC, intra- and interobserver class correlation coefficients.\n\nThe ICC value is the ratio of the between-subject variance to the sum of the between-subject variance and the within-subject variance. The ICC values can be considered as poor (less than 0.40), fair (0.40–0.59), good (0.60–0.74), and excellent (0.75-1.00).\n\n\nDiscussion\n\nOur study shows that there was no significant difference in cardiac inotropic performance evaluated by echocardiography between mice under very light isoflurane anesthesia (HR 475–550 bpm) and awake mice (HR>575 bpm), but that myocardial depression occurred if the anesthesia was too deep (<400 bpm). Echocardiographic evaluation of cardiac function in mice is carried out, most of the times, with the animals under anesthesia, which may alter cardiac function and thereby confound interpretation of the data. Thus, to assess cardiac morphology and function in small animals, it is critical to develop the best technique for accurate and repeated measurements.\n\nPerforming echocardiography under very light anesthesia presents several advantages compared to using awake mice. First, anesthesia permits technically excellent echocardiography by avoiding movement from the mice. Second, as the mouse is under light anesthesia, measurements are not affected by stress, which can alter HR values.\n\nAlthough HR has been suggested to affect echocardiographic measurements2,9,14,19, to our knowledge, no studies have compared the response of echocardiographic measurements between very light anesthesia and awake mice. Roth et al.14 assessed the reproducibility of echocardiographic parameters at several time points 12 days after isoflurane anesthesia, and they found that isoflurane anesthesia provided very good reproducibility on HR, FS and end-diastolic dimensions, compared to intraperitoneal tribromoethanol, ketamine/midazolam or ketamine-xylazine.\n\nNon-invasive assessment of cardiac performance using echocardiography allows for serial evaluations of both function and morphologic parameters. Whether these studies should be performed in conscious restrained animals or under anesthesia remains uncertain. Studies in conscious restrained animals, even after training sessions to prevent bradycardia and make the animals familiar with the procedure, have often reported cardiac function parameters that are significantly higher (HR between 600 and 700 bpm and FS% between 55 to 65)7,19–21 than those quoted for unrestrained animals with telemetry (HR=500–600 bpm, FS% = 35–50)21–24, suggesting sympathetic activation. This was also supported by the fact that responses to parasympathetic blockade with atropine or to administration of isoproterenol were not observed7. Moreover these values were normalized and a full response to isoproterenol was restored when midazolam, a benzodiazepine with little cardiodepressant effect, was administered25. A similar increase in sympathetic discharge was observed in animals with implanted telemeters when subjected to restraint26. In addition, despite animal training and manual restraint, adequate Doppler measurement may be difficult to obtain5. These data suggest that the response to restraint may induce such sympathetic activation that echocardiography functions as a stress test rather than providing an assessment of baseline status. As a matter of fact some of the reported values were comparable to those obtained during sub-maximal exercise (HR> 650 bpm)21.\n\nOn the other hand, previous echocardiographic studies in anesthetized normal mice using various regimens of inhalation and injectable anesthetics at different doses have reported a wide range of LV dimensions (LVEDd 3.1–4.1 mm, FS% 33–58%, HR 250–600, SV 20–50 μl and CO 8–30 ml/min)20,25,27. Some of these have reported heart rates corresponding to 30–40 bpm in humans when echocardiography was performed under anesthesia, suggesting profound hemodynamic depression28. On the basis of our data, we believe that we should use carefully titrated inhalational isofluorane anesthesia in order to minimize both the hemodynamic effect of deep anesthesia and the hemodynamic effect of sympathetic activation in awake mice. The anesthetic regimen is extremely important for improving applicability of animal models to clinical settings, especially in models of shock, in which hemodynamic perturbations are paramount.\n\nOur study did have some limitations. First, the timing of echocardiographic measurements after anesthesia was not studied. Nevertheless, Wu et al.29 recently found that when echocardiographic measurements were performed in mice with a HR between 475 and 525 bpm) after anesthesia, similar echocardiographic parameters could be obtained either a short or long time after anesthesia. As such, timing of the studies should not greatly affect the results. Second, echocardiography was performed on healthy mice and not on mice with pathological conditions. One might imagine, however, that hemodynamics in mice with cardiac disease or shock would be affected even more by anesthesia than healthy mice. Therefore, particular attention should be paid to the anesthesia regimen in mice under pathological conditions using carefully titrated inhalational anesthesia with isoflurane.\n\n\nConclusion\n\nIn conclusion, minimizing the hemodynamic effects of anesthesia is extremely important for improving the applicability of animal models to the clinical setting, especially in models of shock, in which hemodynamic perturbations are significant. We have shown that carefully titrated inhalational anesthesia with isoflurane allowed for echocardiography with minimal perturbation of hemodynamics. This anesthesia regimen allows for application in the study of cardiac function in murine models.\n\n\nData availability\n\nfigshare: Inter and intra-observer reproducibility data for effect of anesthesia on murine cardiac function, doi: http://dx.doi.org/10.6084/m9.figshare.110278030", "appendix": "Author contributions\n\n\n\nFabien A. Picard performed data acquisition and analysis, interpreted the data and drafted the manuscript.\n\nFrançois L. Depret performed data acquisition and analysis, interpreted the data and drafted the manuscript.\n\nSergio Zanotti-Cavazzoni, performed the analysis, interpreted the data and contributed to drafting the manuscript.\n\nSteven M Hollenberg performed the analysis, interpreted the data and contributed to drafting the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nF.D. was supported by a grant from Assistance Publique - Hôpitaux de Paris (Année Recherche).\n\n\nReferences\n\nFox J, Barthold S, Davison M, et al.: The mouse in Biomedical Research. Volume 1–4, Second Edition (American College of Laboratory Animal Medicine). Burlington, MA: Elsevier; 2007. Reference Source\n\nKawahara Y, Tanonaka K, Daicho T, et al.: Preferable anesthetic conditions for echocardiographic determination of murine cardiac function. J Pharmacol Sci. 2005; 99(1): 95–104. PubMed Abstract \| Publisher Full Text\n\nPlante E, Lachance D, Roussel E, et al.: Impact of anesthesia on echocardiographic evaluation of systolic and diastolic function in rats. J Am Soc Echocardiogr. 2006; 19(12): 1520–1525. PubMed Abstract \| Publisher Full Text\n\nPollick C, Hale SL, Kloner RA: Echocardiographic and cardiac Doppler assessment of mice. J Am Soc Echocardiogr. 1995; 8(5 Pt 1): 602–610. PubMed Abstract \| Publisher Full Text\n\nRottman JN, Ni G, Brown M: Echocardiographic evaluation of ventricular function in mice. Echocardiography. 2007; 24(1): 83–89. PubMed Abstract \| Publisher Full Text\n\nTanaka N, Dalton N, Mao L, et al.: Transthoracic echocardiography in models of cardiac disease in the mouse. Circulation. 1996; 94(5): 1109–1117. PubMed Abstract \| Publisher Full Text\n\nYang XP, Liu YH, Rhaleb NE, et al.: Echocardiographic assessment of cardiac function in conscious and anesthetized mice. Am J Physiol. 1999; 277(5 Pt 2): H1967–H1974. PubMed Abstract\n\nRittirsch D, Huber-Lang MS, Flierl MA, et al.: Immunodesign of experimental sepsis by cecal ligation and puncture. Nat Protoc. 2009; 4(1): 31–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHart CY, Burnett JC Jr, Redfield MM: Effects of avertin versus xylazine-ketamine anesthesia on cardiac function in normal mice. Am J Physiol Heart Circ Physiol. 2001; 281(5): H1938–45. PubMed Abstract\n\nMatsuda Y, Ohsaka K, Yamamoto H, et al.: Comparison of newly developed inhalation anesthesia system and intraperitoneal anesthesia on the hemodynamic state in mice. Biol Pharm Bull. 2007; 30(9): 1716–1720. PubMed Abstract \| Publisher Full Text\n\nCollins KA, Korcarz CE, Lang RM: Use of echocardiography for the phenotypic assessment of genetically altered mice. Physiol Genomics. 2003; 13(3): 227–239. PubMed Abstract\n\nButterfield NN, Graf P, Ries CR, et al.: The effect of repeated isoflurane anesthesia on spatial and psychomotor performance in young and aged mice. Anesth Analg. 2004; 98(5): 1305–1311, table of contents. PubMed Abstract \| Publisher Full Text\n\nSzczesny G, Veihelmann A, Massberg S, et al.: Long-term anaesthesia using inhalatory isoflurane in different strains of mice-the haemodynamic effects. Lab Anim. 2004; 38(1): 64–69. PubMed Abstract \| Publisher Full Text\n\nRoth DM, Swaney JS, Dalton ND, et al.: Impact of anesthesia on cardiac function during echocardiography in mice. Am J Physiol Heart Circ Physiol. 2002; 282(6): H2134–2140. PubMed Abstract\n\nZanotti-Cavazzoni SL, Guglielmi M, Parrillo JE, et al.: Fluid resuscitation influences cardiovascular performance and mortality in a murine model of sepsis. Intensive Care Med. 2009; 35(4): 748–754. PubMed Abstract \| Publisher Full Text\n\nOkajima K, Abe Y, Fujimoto K, et al.: Comparative study of high-resolution microimaging with 30-MHz scanner for evaluating cardiac function in mice. J Am Soc Echocardiogr. 2007; 20(10): 1203–1210. PubMed Abstract \| Publisher Full Text\n\nZhang Y, Takagawa J, Sievers RE, et al.: Validation of the wall motion score and myocardial performance indexes as novel techniques to assess cardiac function in mice after myocardial infarction. Am J Physiol Heart Circ Physiol. 2007; 292(2): H1187–1192. PubMed Abstract \| Publisher Full Text\n\nLachin JM: The role of measurement reliability in clinical trials. Clin Trials. 2004; 1(6): 553–566. PubMed Abstract \| Publisher Full Text\n\nJanssen BJ, De Celle T, Debets JJ, et al.: Effects of anesthetics on systemic hemodynamics in mice. Am J Physiol Heart Circ Physiol. 2004; 287(4): H1618–H1624. PubMed Abstract \| Publisher Full Text\n\nRottman JN, Ni G, Khoo M, et al.: Temporal changes in ventricular function assessed echocardiographically in conscious and anesthetized mice. J Am Soc Echocardiogr. 2003; 16(11): 1150–1157. PubMed Abstract \| Publisher Full Text\n\nDesai KH, Sato R, Schauble E, et al.: Cardiovascular indexes in the mouse at rest and with exercise: new tools to study models of cardiac disease. Am J Physiol. 1997; 272(2 Pt 2): H1053–1061. PubMed Abstract\n\nJanssen BJ, Smits JF: Autonomic control of blood pressure in mice: basic physiology and effects of genetic modification. Am J Physiol Regul Integr Comp Physiol. 2002; 282(6): R1545–R1564. PubMed Abstract\n\nButz GM, Davisson RL: Long-term telemetric measurement of cardiovascular parameters in awake mice: a physiological genomics tool. Physiol Genomics. 2001; 5(2): 89–97. PubMed Abstract\n\nUechi M, Asai K, Osaka M, et al.: Depressed heart rate variability and arterial baroreflex in conscious transgenic mice with overexpression of cardiac Gsalpha. Circ Res. 1998; 82(4): 416–423. PubMed Abstract \| Publisher Full Text\n\nIshizaka S, Sievers RE, Zhu BQ, et al.: New technique for measurement of left ventricular pressure in conscious mice. Am J Physiol Heart Circ Physiol. 2004; 286(3): H1208–1215. PubMed Abstract \| Publisher Full Text\n\nKramer K, van Acker SA, Voss HP, et al.: Use of telemetry to record electrocardiogram and heart rate in freely moving mice. J Pharmacol Toxicol Methods. 1993; 30(4): 209–215. PubMed Abstract \| Publisher Full Text\n\nTakuma S, Suehiro K, Cardinale C, et al.: Anesthetic inhibition in ischemic and nonischemic murine heart: comparison with conscious echocardiographic approach. Am J Physiol Heart Circ Physiol. 2001; 280(5): H2364–H2370. PubMed Abstract\n\nKass DA, Hare JM, Georgakopoulos D: Murine cardiac function: a cautionary tail. Circ Res. 1998; 82(4): 519–522. PubMed Abstract \| Publisher Full Text\n\nWu J, Bu L, Gong H, et al.: Effects of heart rate and anesthetic timing on high-resolution echocardiographic assessment under isoflurane anesthesia in mice. J Ultrasound Med. 2010; 29(12): 1771–1778. PubMed Abstract\n\nPicard FA, Depret FL, Zanotti Cavazzoni SL, et al.: Inter and intra-observer reproducibility data for effect of anesthesia on murine cardiac function. Figshare. 2014. Data Source" }	[ { "id": "6357", "date": "07 Oct 2014", "name": "Abdallah Fayssoil", "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 interesting and original research that evaluates the effect of anesthesia level on murine cardiac function, using a non-invasive approach. The data are interesting and will be helpful to the scientific community.", "responses": [] }, { "id": "6842", "date": "15 Dec 2014", "name": "Matthias Riess", "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 \"Effect of anesthesia level on murine cardiac function“ F Picard et al. anesthetized 12 mice three times at different concentrations of the volatile anesthetic isoflurane to test the hypothesis that isoflurane has a dose-dependent negative inotrope and chronotrope effect on myocardial function. Overall, this is an interesting and useful study to help guide basic science investigators when conducting echocardiography in cardiovascular studies in mice. Several questions/concerns, however, would need to be addressed to improve the validity of the data and the value for F1000Research’s readers: Please integrate more details on the (revised) statistical analysis into the abstracts as well. Introduction, 2nd paragraph: most importantly, echocardiograpy is non-invasive. Introduction, 2nd paragraph: I do not agree with the statement that “… physiological mechanisms in mice are closely related to those in humans.” With a heart rate in the 500s and a correspondingly lower stroke volume this is hardly the case. In addition, the body mass to surface ratio and therefore metabolism are different. Please revise accordingly. Introduction, 3rd paragraph: “In fact, heart rate can be used as a proxy for depth of anesthesia.” While this may be true for the murine model at hand, depth of anesthesia in general cannot be assessed by heart rate only and strongly depends on the anesthetic used (volatile vs intravenous, cardiovascular side effects vary greatly among drugs, hemodynamic status, acutely and chronically administered concurrent medications, etc). Please revise. Material and Methods, 1st paragraph: please use the current version of the Guide. Material and Methods, 2nd paragraph (Experimental Protocols), and study design in general: the authors seem to exchange dependent and independent variables in their experiments. Either isoflurane is given at a predetermined end tidal concentration as an independent variable with a negligible standard deviation in its concentration in each group, then heart rate is a truly dependent variable. If, however, isoflurane is titrated to achieve a certain heart rate, the latter turns into an independent variable while isoflurane becomes the dependent variable. Either way, isoflurane concentrations and heart rates in each group need to be shown as mean plus/minus standard deviation in case of normal distribution or median and quartiles if not normally distributed. Please revise, including table 1. Material and Methods, 4th paragraph: what were the animals’ temperatures? Deviations in temperature may have a significant effect on heart rate independent of the isoflurane concentration. Material and Methods, 4th paragraph: was an end tidal isoflurane concentration measured or are the concentrations provided from the vaporizer setting? If so, what fresh gas flow was used? Statistical analysis: more than one comparison requires an ANOVA with post-hoc comparisons, not repeated t-tests to avoid a type I error. For non-parametric testing see below. Table 1: The statistical symbols are largely redundant, e.g. if deep is different to light, light does not have to be shown again to be different to deep etc. Please simplify. Figure 1: The nature of the box plots suggests that most of the data are not normally distributed. Therefore, non-parametric tests and data presentation are necessary. Please revise. As laid out before, due to a switch between strictly independent and dependent variables there appears to be a greater than necessary heterogeneity in the isoflurane concentrations in any of the three groups. Therefore, please plot the outcome data presented in panels A through D of figure 1 as scatter plots with the isoflurane concentrations on the x-axis and the chosen outcome on the y-axis and conduct regression and correlation analyses to emphasize the take home points of this study.", "responses": [] } ]	1	https://f1000research.com/articles/3-165
https://f1000research.com/articles/3-127/v2	18 Jun 14	{ "type": "Research Article", "title": "Validation of the peroneal nerve test to diagnose critical illness polyneuropathy and myopathy in the intensive care unit: the multicentre Italian CRIMYNE-2 diagnostic accuracy study", "authors": [ "Nicola Latronico", "Giovanni Nattino", "Bruno Guarneri", "Nazzareno Fagoni", "Aldo Amantini", "Guido Bertolini", "and GiVITI Study Investigators", "Giovanni Nattino", "Bruno Guarneri", "Nazzareno Fagoni", "Aldo Amantini", "Guido Bertolini" ], "abstract": "Objectives: To evaluate the accuracy of the peroneal nerve test (PENT) in the diagnosis of critical illness polyneuropathy (CIP) and myopathy (CIM) in the intensive care unit (ICU). We hypothesised that abnormal reduction of peroneal compound muscle action potential (CMAP) amplitude predicts CIP/CIM diagnosed using a complete nerve conduction study and electromyography (NCS-EMG) as a reference diagnostic standard.Design: prospective observational study.Setting: Nine Italian ICUs.Patients: One-hundred and twenty-one adult (≥18 years) neurologic (106) and non-neurologic (15) critically ill patients with an ICU stay of at least 3 days.Interventions: None.Measurements and main results: Patients underwent PENT and NCS-EMG testing on the same day conducted by two independent clinicians who were blind to the results of the other test. Cases were considered as true negative if both NCS-EMG and PENT measurements were normal. Cases were considered as true positive if the PENT result was abnormal and NCS-EMG showed symmetric abnormal findings, independently from the specific diagnosis by NCS-EMG (CIP, CIM, or combined CIP and CIM). All data were centrally reviewed and diagnoses were evaluated for consistency with predefined electrophysiological diagnostic criteria for CIP/CIM.During the study period, 342 patients were evaluated, 124 (36.3%) were enrolled and 121 individuals with no protocol violation were studied. Sensitivity and specificity of PENT were 100% (95% CI 96.1-100.0) and 85.2% (95% CI 66.3-95.8). Of 23 patients with normal results, all presented normal values on both tests with no false negative results. Of 97 patients with abnormal results, 93 had abnormal values on both tests (true positive), whereas four with abnormal findings with PENT had only single peroneal nerve neuropathy at complete NCS-EMG (false positive).Conclusions: PENT has 100% sensitivity and high specificity, and can be used to diagnose CIP/CIM in the ICU.", "keywords": [ "Sensitivity", "specificity", "critical illness polyneuropathy", "critical illness myopathy", "muscle weakness", "multiple organ failure" ], "content": "Introduction\n\nCritical illness polyneuropathy (CIP) affects 30% to 50% of the most severely critically ill patients and is the most frequent acute polyneuropathy in the intensive care unit (ICU)1. CIP is often associated with an acute, primary myopathy called critical illness myopathy (CIM), and both conditions occur in patients with multiple organ dysfunctions and failure (MOF). Indeed, CIP and CIM represent the failure of the neuromuscular system in patients with MOF1. CIP classically presents as a sensory-motor axonal polyneuropathy causing difficulty in weaning patients from a ventilator, flaccid limbs, and a possible reduction in deep tendon reflexes. Amplitude reduction of both the compound muscle action potential (CMAP) and the sensory nerve action potential (SNAP) is the predominant electrophysiological finding in CIM and CIP; latency and nerve conduction velocity remain normal or are only slightly decreased. CIM is a primary myopathy with distinctive electrophysiological and morphological findings1. Other features of CIM include increased CMAP duration, normal SNAPs, reduced muscle excitability on direct stimulation and myopathic motor unit potentials on needle electromyography. The clinical features are often the same as for CIP, but sensation, if testable, is normal.\n\nTraditional methods to diagnose CIP and CIM include manual testing of muscle strength using the UK Medical Research Council (MRC) score2 or dynamometry3 to demonstrate severe weakness, electrophysiological tests to explore the function of peripheral nerves and muscles, and muscle biopsy1. Conventional nerve conduction studies (NCS) with measurement of conduction velocity, CMAP and SNAP amplitudes, electromyography (EMG) and other specialised techniques such as direct muscle stimulation or axonal excitability testing may reveal nerve or muscle dysfunction with a high degree of specificity4. However, these techniques require specialised personnel, they are time-consuming and they do not allow diagnosis of small intra-epidermal nerve fiber pathology5. Conventional NCS-EMG may require up to 90 minutes to be completed4. Considering the high prevalence of ICU-acquired neuromuscular disorders, it is unrealistic for conventional NCS-EMG to be used as a large-scale screening tool.\n\nElectrophysiological investigations of peripheral nerves and muscles offer several advantages. They are minimally invasive and easily repeatable, they can be performed at the bedside and the results are immediately available6. Electrophysiological alterations indicating CIP or CIM can also be demonstrated in non-collaborative patients. In comatose patients or in those with persisting sedation or septic encephalopathy who develop severe muscle weakness or paralysis after ICU admission, electrophysiological investigations can be performed to avoid unreasonably pessimistic prognosis by identification of CIP or CIM as the cause7. Electrophysiological alterations are not only essential to establish the diagnosis of CIP and CIM, but they also have an earlier onset than clinical signs or they can be documented at an earlier stage7–11, thus offering the advantage of a timely diagnosis and the opportunity to conduct potentially valuable interventions before structural muscle-nerve alterations become established4. CMAP reduction is an early event which precedes clinical signs: its onset can be abrupt within 24 hours of normal findings10, and it can be observed as early as 48 hours before clinical signs in patients with sepsis12. Moreover, NCS may also be useful to predict hospital mortality9, and short13 and long-term morbidity8,14,15. CMAP after direct muscle stimulation (dmCMAP) may precede the development of ICU-acquired weakness by several days11. Therefore there is a need for a rapid, simple, accurate and valid electrophysiological test to identify CIP and CIM early in ICU.\n\nIn a previous study called CRIMYNE10, we found that a simplified electrophysiological investigation, the peroneal nerve electrophysiological test (PENT), had high sensitivity (100%) and moderate specificity (67%) in identifying patients with a diagnosis of CIP or CIM using complete NCS as the reference diagnostic standard. However, the CRIMYNE study had methodological limitations that precluded the acceptance of PENT as a valid screening test in the ICU. Specifically, patients underwent complete NCS only if the PENT was abnormal, thus precluding the possibility of accurately detecting false negative results. Moreover, NCS assessors were not blind to the results of PENT. According to the STARD Statement for reporting studies of diagnostics accuracy, knowledge of the results of the index test can influence the reading of the reference standard, and vice versa. Such knowledge is likely to increase the agreement between results of the index test and those of the reference standard, leading to inflated measures of diagnostic accuracy16.\n\nThe accuracy of new diagnostic tests should be evaluated before their introduction into clinical practice, because invalid tests may yield exaggerated and biased results, which may trigger their premature adoption in clinical practice and lead to test overuse, increasing patient risks and costs, or to test underuse, missing opportunities to improve health17,18. Test validation involves comparing the new test with a reference standard, defined as the best available method for establishing the presence or absence of the condition of interest17. Therefore, we designed the multicentre CRIMYNE-2 study according to rigorous methodology to provide an unbiased evaluation of the accuracy of PENT in diagnosing electrophysiologically-proven CIP and CIM diagnosed using complete NCS-EMG as the reference standard.\n\n\nMaterials and methods\n\nWe conducted a prospective observational study between April 2010 and June 2012 in nine medical-surgical Italian ICUs joining the Margherita-Prosafe Project, an international research campaign that collects and analyses clinical data on all patients admitted to the participating ICUs19. The Project, promoted by the GiViTI (Gruppo Italiano per la Valutazione degli Interventi in Terapia Intensiva, Italian Group for the Evaluation of Interventions in Intensive Care Medicine) is based on an electronic Case Report Form (eCRF), which was extended to collect the data of interest for the CRIMYNE-2 study. Two databases were used: Margherita-Due, supported until December 2010, and Prosafe, released on January 2011. Both databases were organized with a core data set and a supplemental module designed for the CRIMYNE-2 research project.\n\nThe core data of the Margherita-Prosafe Project included information related to the patients' condition on admission (demographics, admission diagnoses, comorbidities, surgical status), severity-of-illness scoring systems (Simplified Acute Physiology Score [SAPS II], Glasgow Coma Scale [GCS]); procedures, organ failures and complications arising during the ICU; ICU and hospital outcome. The CRIMYNE-2 module included the electrophysiological variables needed to diagnose CIP and CIM, as well as the final diagnosis concluded by the clinical neurophysiologist (see “Possible diagnoses”).\n\nInclusion criteria were age ≥ 18 years and ICU stay ≥ 3 days, because this patient population is at increased risk of developing CIP and CIM1,20.\n\nExclusion criteria were a previous history of neuromuscular disorders or chronic conditions associated with neuromuscular disorders (i.e., diabetes, renal failure, chronic alcohol abuse), use of neurotoxic or myotoxic drugs, evidence of altered neuromuscular transmission at repetitive stimulation test either caused by neuromuscular blocking agents or disease, lower limb disorders precluding NCS and EMG (for example edema, fractures, amputation, plaster casts), and terminal conditions. Sedation was not an exclusion criterion, and, if present, was not discontinued nor reduced prior to testing.\n\nIn each ICU, the inclusion and exclusion criteria were verified daily by the intensivist in charge for the CRIMYNE-2 study. Patients enrolled were subjected to the index test (PENT) and the reference diagnostic standard (NCS-EMG) on the same day by two independent clinicians who were blind to the results of the other test. The research team was allowed to exclude patients if another patient in the same ICU was being concomitantly studied.\n\n\nEthics statement\n\nThe study was approved by the Ethics Committee of each participating centre (protocol number 554/2010). Detailed written information was provided to the patients and family members about the study protocol, the scope of research, and the safety of electrophysiological investigations. Written informed consent was obtained from the patient where appropriate. In case of altered consciousness, the Ethics Committees waived the requirement for consent, as in Italy relatives are not regarded as legal representatives of the patient in the absence of a formal designation21. Written informed consent was requested from all surviving patients as soon as they regained their mental competency. All investigations were conducted according to the principles expressed in the Declaration of Helsinki.\n\nAll tests were performed and interpreted by a board-certified technician in clinical neurophysiology with adequate clinical experience.\n\nPENT started with the measurement of peroneal nerve CMAP amplitude in one leg (Figure 1). The CMAP was recorded using a pair of surface electrodes: the active electrode was placed on the belly of the extensor digitorum brevis muscle and the indifferent electrode on the distal tendon of the recorded muscle. The peroneal nerve was stimulated over the dorsum of the foot near the ankle, at 7 centimeters from the recording electrodes. The stimulus intensity was gradually increased until the maximal CMAP was obtained. The CMAP amplitude was measured as the maximum voltage difference between the negative and positive peaks (\"peak to peak\").\n\nA normal compound muscle action potential (CMAP) amplitude is shown.\n\nIf the PENT was normal, the contra-lateral peroneal nerve was investigated. A normal condition was established if the CMAP amplitudes of both peroneal nerves were normal. An abnormal condition was identified if the peroneal nerve CMAP amplitude was reduced below the normal limits of the participating centre in at least one leg.\n\nNCS-EMG included conduction studies of motor and sensory nerves, and needle EMG, which were performed by board-certified clinical neurophysiologists.\n\nMotor nerves included22–24: a) the ulnar nerves with recording from the abductor digiti minimi muscle and stimulation at wrist and above the elbow bilaterally; b) the peroneal nerves with recording from the extensor digitorum brevis muscle and stimulation on the dorsum of the foot and below the fibular head bilaterally, and c) the posterior tibial nerves with recording from the abductor hallucis muscle and stimulation at the medial malleolus and popliteal fossa bilaterally. Recorded parameters included the conduction velocity and distal latency, the amplitude of the CMAP and of the recurrent responses after supra-maximal intensity distal stimulation (F response to stimulation of the posterior tibial nerve and of the ulnar nerve).\n\nSensory nerves included: a) antidromic conduction studies of the ulnar nerves with stimulation at wrist level and recording of 5th finger bilaterally, and b) antidromic conduction study of the sural nerves with stimulation at the lower third of the leg and recording at the lateral malleolus bilaterally.\n\nIf a neuromuscular transmission defect was suspected, as in case of recent use of neuromuscular blocking agents, low frequency (3Hz) repetitive stimulation studies were performed with stimulation of the right ulnar nerve and recording from abductor digiti minimi and stimulation at wrist. In the case of decremental response on repetitive nerve stimulation, the patient was excluded from the study.\n\nEMG was recorded using a coaxial needle electrode in proximal and distal muscles of upper (brachial biceps, abductor digiti minimi) and lower limbs (quadriceps femori, tibialis anterior). Abnormal spontaneous muscle activity in forms of fibrillation potentials and positive sharp waves was classified in 4 levels (0, 1+, 2+, 3+). Motor units were evaluated in terms of amplitude, duration and morphology and recruitment patterns in patients able to activate their muscles at their own volition25,26.\n\nPossible diagnoses based on complete NCS-EMG were: CIP, CIM, combined CIP and CIM, undetermined, or normal findings. The diagnostic criteria used conformed to accepted standards and have been described in detail elsewhere1,10,24.\n\nWe did not evaluate the muscle strength clinically, hence the diagnoses were defined as probable1. Probable CIP was established if NCS showed a reduction in the amplitude of CMAP and SNAP below the normal value of the laboratory, with normal or mildly reduced nerve conduction velocity and normal neurotransmission1.\n\nProbable CIM was established in collaborative patients if the CMAP was abnormally reduced below the normal value of the laboratory, the SNAP was normal, and needle EMG demonstrated low-amplitude motor unit potentials with short duration, and early or normal full recruitment, with or without fibrillation potentials1.\n\nIn non-collaborative patients, where the differential diagnosis between CIP and CIM could not be established using conventional NCS and EMG4,27, centres were not required to use specialised neurophysiological techniques such as direct muscle stimulation, and abnormal findings were classified as undetermined.\n\nCases were considered as true negative if both ENG-EMG and PENT assessments were normal. Cases were considered as true positives if the PENT test was abnormal and ENG-EMG showed symmetric abnormal findings, independently from the specific diagnosis at NCS-EMG (CIP, CIM, or combined CIP and CIM). Patients with an abnormal finding at PENT assessment and normal findings or non-symmetrical neuropathy at ENG-EMG (i.e. mononeuropathy or multineuropathy) were considered as false positives.\n\n\nData quality control\n\nAll data were centrally reviewed by four of the research team (NL, BG, GN, GB), and diagnoses were evaluated for consistency with predefined electrophysiological diagnostic criteria of CIP and CIM. In the case of discordant results, the centres were contacted and diagnoses were discussed with local personnel until a consensus was reached. If needed, supplemental electrophysiological material was requested from the participating centre and analysed.\n\nThe study results are reported according to the Standards for Reporting of Diagnostic Accuracy (STARD)17.\n\nWe estimated a prevalence of CIP or CIM of 0.3, a test sensitivity of 90%, and test specificity of 65% based on the results of the CRIMYNE study10. We set a clinically acceptable precision at 10% for estimates of both sensitivity (true positive rate: 85% to 95%) and specificity (true negative rate: 60%–70%), and calculated that 125 patients would be needed to achieve such a precision28.\n\nWe described continuous variables as means and standard deviations (SD) or medians and interquartile range (IQR), and categorical variables as counts and percentage.\n\nThe 95% confidence intervals (95% CI) for sensitivity and specificity were calculated according to the binomial distribution.\n\n\nResults\n\nDuring the study period, 342 patients were evaluated and 124 (36.3%) were enrolled (Figure 2).\n\nDiscordant results after central revision were documented in five cases. Two cases, classified by the local clinical neurophysiologist as normal despite minimally reduced ulnar CMAP amplitudes, were confirmed as normal at central revision after upgrading the outdated limits of normality used centrally. One case classified locally as abnormal because of minimally reduced proximal peroneal CMAP amplitude (recorded above the fibular head) and normal distal peroneal CMAP (below the fibular head) was classified as normal after central revision, according to the study protocol. One case initially misclassified as having normal PENT because of erroneous transcription of the data was reclassified to abnormal PENT after re-examination of the original electrophysiological data. One case was eventually classified as showing inconclusive results. This patient had normal findings at PENT and abnormal findings at complete NCS-EMG performed 24 hours later. Because this was a protocol violation and CMAP reduction may develop rapidly, within 24 hours of normal findings10, we excluded the patient from the analysis.\n\nOf the 124 patients enrolled, 121 (98%) were investigated with both the index test (PENT) and the reference standard (complete NCS-EMG) with no protocol violation (Figure 2). The general characteristics of the study population are presented in Table 1. Acute neurologic patients were the main group of patients, with head trauma (35 patients), spontaneous intracranial haemorrhage (25) and aneurysmal subarachnoid haemorrhage (21) being the most frequent causes of admission. Patients had a severe clinical condition, with high prevalence of sepsis, single organ failure (25%) or MOF (62%), and prolonged duration of mechanical ventilation and ICU stay, particularly in patients with abnormal NCS-EMG. Hospital mortality was higher in patients with abnormal NCS-EMG results than in patients with a normal NCS-EMG (Table 1).\n\nGCS: Glasgow Coma Scale; SAPS: Simplified Acute Physiology Score. NCS: nerve conduction study-electromyography.\n\n* Percentages may sum to more than 100% because patients may be classified in more than one category.\n\nOf 23 patients with normal NCS-EMG, all also had normal findings with PENT with no false negative results (Figure 2). Of 97 patients with abnormal ENG-EMG, four had peroneal nerve mononeuropathy at complete NCS-EMG and abnormal findings at PENT, and were classified as false positives. The sensitivity of the PENT was 100% (95% CI 96.1-100.0) and the specificity 85.2% (95% CI 66.3-95.8). Sensitivity was comparable to that found in the CRIMYNE study (100%). Specificity was higher in this study compared to the CRIMYNE study, but the difference was not statistically significant (CRIMYNE: 67%, CRIMYNE-2: 85%, p=0.08).\n\nThe electrophysiological diagnosis in the 93 patients with abnormal findings at complete NCS-EMG was CIP in 35 patients (37.6%), combined CIP and CIM in 16 (17.2%) patients and undetermined in 42 (45.2%) patients.\n\nElectrophysiological investigations were performed on median ICU day 9 (IQR 5-16). The median time needed to perform PENT and the complete NCS-EMG was 10 minutes (IQR 8.0-10.5) and 50 minutes (40–60). In 85 cases (70.2%), the peroneal nerve examined first was abnormal, thus eliminating the need for contralateral testing.\n\n\nDiscussion\n\nIn this multicentre diagnostic accuracy study, we found that the PENT had 100% sensitivity and high specificity in diagnosing CIP or CIM, confirming the preliminary results of the previous CRIMYNE study10. The time needed to complete the test was 10 minutes, which was considerably shorter than the time needed to complete NCS-EMG (50 minutes).\n\nConventional NCS and EMG have never gained popularity in the ICU because they are time-consuming, not readily available, expensive, and technically challenging as they require specialised personnel. Therefore, our results may be important in promoting a wider use of electrophysiological investigations in the ICU. As a validated, high-sensitivity, minimally invasive, non-volitional and quick diagnostic test, PENT can accurately exclude CIP or CIM if the result is normal, and can be proposed for the assessment and monitoring of the neuromuscular function in the early stages of critical illnesses.\n\nAbnormal PENT cannot discriminate between CIP, CIM or combined CIP and CIM; in fact, in our study, diagnosis was undetermined in almost half of patients. Because precise definition of pathology can be relevant for predicting the recovery of muscle strength13,14, PENT can be used as a screening tool to select patients deserving further investigations. Moreover, “false positive patients” are those with peroneal nerve mononeuropathy who are worthy of medical attention. Involvement of the peroneal nerve is invariably detected in patients with CIP or CIM, possibly because it is the longest nerve of the body and, hence, it is mostly vulnerable to the energy deficit caused by tissue ischemia or dysoxia7,10. In fact, the biosynthesis takes place in the neuronal cell body and then the structural components are moved into the axons and transported to their final destination to generate the action potential and to maintain axonal integrity, a process that requires considerable amount of energy7,10. This might explain the reason why reduced peroneal nerve CMAP is highly sensitive compared to other tests. Pragmatically, measurements of peroneal CMAP could be implemented to diagnose all patients with ICU-acquired neuromuscular disorders, either CIP or CIM, while other tests, such as measurement of dmCMAP, could be used to differentiate CIM from CIP29. Recent studies suggest that the duration of the CMAP can be prolonged in patients with CIM30.\n\nFive single-centre studies have used simplified electrophysiological tests of neuromuscular function in the ICU10,11,31–33. In two randomized controlled trials on intensive insulin treatment31,32, the diagnosis of CIP was suggested by the presence of abundant spontaneous activity in the form of positive sharp waves and fibrillation potentials on EMG34. No reference diagnostic standard was used, thus the sensitivity and specificity of the test are unknown. In the CRIMYNE study10, we used the same index test and reference diagnostic standard of the CRIMYNE-2 study, but comparison was unblinded and was limited to patients with abnormal findings at the index test, which may have inflated the test sensitivity. In a prospective observational study in surgical ICU patients11, the sensitivity and specificity of dmCMAP in predicting the development of ICU-acquired weakness were 83.3% and 88.8%, respectively. In a more recent study33, sensitivity and specificity of combined peroneal nerve CMAP and sural nerve SNAP evaluations were 100% and 81%, respectively. However, in both studies all electrophysiological measurements were performed by the same investigators, and no blind assessment of the reference diagnostic test was implemented35.\n\nAmong non-volitional non-electrophysiological tests, ultrasound-guided assessment of muscle mass can reliably detect early muscle changes. In a recent prospective cohort study in 63 critically ill patients who were developing acute muscle wasting, the rectus femoris cross-sectional area decreased significantly from days 1 to 7 by 12.5%36. Simultaneous measurement of muscle strength and muscle mass has been suggested as a critical step in all future studies investigating ICU-acquired weakness37. However, assessing muscle strength may be difficult in the early stage of acute disease, and the use of PENT could be of value in assessing the generation of the action potential, which is an essential pre-requisite for muscle contraction.\n\nThe prevalence of electrophysiological abnormalities was 76%, which is higher than reported in a systematic review [46% (95% C.I. 43–49%)]20. We enrolled patients with high incidence of sepsis and MOF, and prolonged mechanical ventilation5. It is therefore not surprising that this patient group also had a high incidence of ICU-acquired neuromuscular disorders that are currently viewed not as isolated events, but rather as an integral part of the process leading to MOF1. In two recent studies, severe muscle weakness was documented in 74% of patients38 and electrophysiological abnormalities in 87%11. In an unselected population of critically ill children, the occurrence of generalised muscle weakness was substantially lower (1.7%)39, reflecting the need for future studies to include much larger samples of unselected ICU patients to evaluate the true incidence of ICU-acquired neuromuscular disorders.\n\nSeveral limitations of the current study are worth discussing. First, we did not test muscle strength clinically, and therefore, we could not evaluate the proportion of patients with pure electrophysiological alterations that developed definite CIP or CIM. Current recommendations suggest that patients with a MRC sum-score of less than 48 or reduced handgrip dynamometry should undergo physical rehabilitation without any further testing1,40. NCS and EMG testing should be reserved for those patients not improving despite receiving such treatment. MRC and dynamometry can be easily implemented in the ICU, and inter-observer reliability is good. However, volitional tests require the patient’s collaboration, hence an inability to perform the test or the recording of low values may occur as a result of coma, delirium, sedation, injury, or simply poor patient motivation or attention. Recent studies suggest that manual muscle testing can be insufficient for the early detection of ICU-acquired neuromuscular disorders in most patients41. In fact, ascertainment of ICU-acquired weakness at ICU day 7 post-awakening, as originally defined2, can be of limited value because many patients have been already discharged from the ICU at that time, reflecting the change in clinical ICU practice towards earlier discharge38. Moreover, inter-observer agreement on MRC sum-score assessment is only moderate in patients with MRC scores that are lower than 48, probably because the ability to perform the volitional MRC sum-score among the most severe patients is variable38,42. PENT is not dependent upon patients’ collaboration and could be performed at an early stage to detect initial functional derangement and to use the results to start investigational treatments with the aim of interrupting pathological mechanisms at their onset43. Several studies have demonstrated that electrophysiological alterations are followed by muscle weakness, but they included a small number of selected ICU patients7–11. PENT is a quick test and could be used in large ICU populations.\n\nSecond, the large majority of patients were critically ill neurologic patients, which may limit the generalisability of the results to other ICU populations. However, development of CIP and CIM is independent from the cause of ICU admission, and it rather depends on MOF, which is also a common complication in acute neurologic patients44, as it is confirmed in this study.\n\nLastly, we excluded patients with diabetes, who may represent up to 16% percent of patients admitted to the ICU45. Peripheral neuropathy is a common complication in diabetic patients46. However, generalised symmetric polyneuropathy typically involves sensory nerves with mild reduction in distal sensory response amplitudes (e.g., those of the sural nerves)47. In contrast, motor response amplitudes, as those tested with peroneal nerve evaluation, are generally preserved and decrease only in more advanced disease.\n\n\nConclusions\n\nTo the authors’ knowledge, this is the first time that a simplified electrophysiological test to assess the peripheral nerve and muscle function has been evaluated in an ICU population using a rigorous methodology based on independent and blinded comparison with a reference standard.\n\nMeasurement of CMAP of the peroneal nerve showed 100% sensitivity and high specificity in diagnosing probable CIP or CIM, and did not require patients’ collaboration. Potential useful applications of the test can be at the early ICU stage, when volitional tests can be rarely performed. At a later stage, before discharge from ICU or acute-care hospital, a normal test excludes CIP or CIM and the need for further electrophysiological investigations. If patients present abnormal values, they might have CIP, CIM or focal peripheral nerve complications such as peroneal entrapment neuropathy that warrant neurological consultation.\n\nThe application of this test to wider populations of ICU patients might allow a more precise estimation of the true incidence of ICU-acquired neuromuscular disorders. With adequately powered observational cohort studies, future studies might be able to evaluate the relationship between electrophysiologically-proven CIP or CIM and reduced muscle strength and mass in the ICU36,37, as well the relationship between these acute changes and persistent muscle weakness or persistent physical dysfunction at long-term follow-up42.\n\n\nData availability\n\nfigshare: Data set for the CRIMYNE-2 study on the validation of peroneal nerve test to diagnose polyneuropathy and myopathy in 121 patients, doi: http://dx.doi.org/10.6084/m9.figshare.102150648\n\n\nConsent\n\nDetailed written information was provided to the patients and family members about the study protocol, the scope of research, and the safety of electrophysiological investigations. Written informed consent was obtained from the patient where appropriate. In case of altered consciousness, the Ethics Committees waived the requirement for consent, as in Italy relatives are not regarded as legal representatives of the patient in the absence of a formal designation21. Written informed consent was requested from all surviving patients as soon as they regained their mental competency.", "appendix": "Author contributions\n\n\n\nAll authors made a substantial contribution to the study design and methods.\n\nNicola Latronico conceived the idea of the study. Nicola Latronico, Guido Bertolini, Bruno Guarneri and Aldo Amantini designed the protocol.\n\nGuido Bertolini and Giovanni Nattino performed the statistical analyses.\n\nGuido Bertolini and Giovanni Nattino were responsible for database handling and data quality control.\n\nBruno Guarneri, Nicola Latronico, Guido Bertolini and Giovanni Nattino were responsible for the central revision of electrophysiological diagnoses.\n\nNicola Latronico and Nazzareno Fagoni were responsible for the clinical investigations of the study.\n\nNicola Latronico drafted the manuscript and all authors critically revised it for important intellectual content. All authors read and approved the final manuscript.\n\nGiViTI group was responsible for collecting data.\n\nThe full list of collaborators (GiViTI group) is as follows:\n\nSteering committee: Luisa Antonini, MD (University of Brescia, Italy); Matthias Eikermann, MD (Harvard University, Boston, USA); Tarek Sharshar, MD, PhD (University of Versailles, France) and Robert Stevens, MD, PhD (John Hopkins University Schools of Medicine, Baltimore, USA).\n\nGiVITI study investigators: Bruno Guarneri, MD, Lucia Salvatoni, technician, Nazzareno Fagoni, MD, Simone Piva, MD, Frank Rasulo, MD and Nicola Latronico, MD (University of Brescia at Spedali Civili, Brescia, Italy); Maurizio Osio, MD, Anna Mangiatordi, technician and Beatrice Borghi, MD (Ospedale Sacco, Milano, Italy); Enrico Bosco, MD, Fabrizio Baldanzi, technician, and Ennio Nascimben, MD (Ospedale Regionale S. Maria dei Battuti Ca' Foncello, Treviso, Italy); Camillo Foresti, MD, Paolo Conte, technician, and Federica Rottoli, MD (Ospedale Papa Giovanni XXIII, Bergamo, Italy); Luisa Motti, MD, Claudio Basile, technician, and Valter Bottari, MD (Ospedale S. Maria Nuova, Reggio Emilia, Italy); Aldo Amantini, MD, Cesarina Cossu, technician, and Manuela Bonizzoli, MD (Ospedale Careggi, Firenze, Italy); Giovanna Squintani, MD, Silvia Romito, MD, and Francesca Casagrande, MD (Ospedale Maggiore Borgo Trento, Verona, Italy); Marcello Romano, MD, Simone Caruso, technician, Baldassare Renda, MD (Ospedale Villa Sofia-CTO, Palermo, Italy); Ennio Iezzi, MD, Luigi Chiacchiari, technician, and Valeria Marinò, MD (Istituto Neurologico Mediterraneo IRCCS Neuromed, Pozzilli, Italy).\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe study was funded with internal funds of the University of Brescia, Italy and the IRCCS-Istituto di Ricerche Farmacologiche Mario Negri, Italy.\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\nLatronico N, Bolton CF: Critical illness polyneuropathy and myopathy: a major cause of muscle weakness and paralysis. Lancet Neurol. 2011; 10(10): 931–41. PubMed Abstract \| Publisher Full Text\n\nDe Jonghe B, Sharshar T, Lefaucheur JP, et al.: Paresis acquired in the intensive care unit: a prospective multicenter study. JAMA. 2002; 288(22): 2859–67. PubMed Abstract \| Publisher Full Text\n\nAli NA, O'Brien JM Jr, Hoffmann SP, et al.: Acquired weakness, handgrip strength, and mortality in critically ill patients. Am J Respir Crit Care Med. 2008; 178(3): 261–8. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Shehu I, Guarneri B: Use of electrophysiologic testing. Crit Care Med. 2009; 37(10 Suppl): S316–20. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Filosto M, Fagoni N, et al.: Small nerve fiber pathology in critical illness. PLoS One. 2013; 8(9): e75696. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolton CF: Neuromuscular manifestations of critical illness. Muscle Nerve. 2005; 32(2): 140–63. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Fenzi F, Recupero D, et al.: Critical illness myopathy and neuropathy. Lancet. 1996; 347(9015): 1579–82. PubMed Abstract \| Publisher Full Text\n\nLeijten FS, Harinck-de Weerd JE, et al.: The role of polyneuropathy in motor convalescence after prolonged mechanical ventilation. JAMA. 1995; 274(15): 1221–5. PubMed Abstract \| Publisher Full Text\n\nKhan J, Harrison TB, Rich MM, et al.: Early development of critical illness myopathy and neuropathy in patients with severe sepsis. Neurology. 2006; 67(8): 1421–5. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Bertolini G, Guarneri B, et al.: Simplified electrophysiological evaluation of peripheral nerves in critically ill patients: the Italian multi-centre CRIMYNE study. Crit Care. 2007; 11(1): R11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeber-Carstens S, Koch S, Spuler S, et al.: Nonexcitable muscle membrane predicts intensive care unit-acquired paresis in mechanically ventilated, sedated patients. Crit Care Med. 2009; 37(9): 2632–7. PubMed Abstract \| Publisher Full Text\n\nTennila A, Salmi T, Pettila V, et al.: Early signs of critical illness polyneuropathy in ICU patients with systemic inflammatory response syndrome or sepsis. Intensive Care Med. 2000; 26(9): 1360–3. PubMed Abstract \| Publisher Full Text\n\nKoch S, Spuler S, Deja M, et al.: Critical illness myopathy is frequent: accompanying neuropathy protracts ICU discharge. J Neurol Neurosurg Psychiatry. 2011; 82(3): 287–93. PubMed Abstract \| Publisher Full Text\n\nGuarneri B, Bertolini G, Latronico N: Long-term outcome in patients with critical illness myopathy or neuropathy: the Italian multicentre CRIMYNE study. J Neurol Neurosurg Psychiatry. 2008; 79(7): 838–41. PubMed Abstract \| Publisher Full Text\n\nKoch S, Wollersheim T, Bierbrauer J, et al.: Long term Recovery in Critical Illness Myopathy is complete, contrary to Polyneuropathy. Muscle Nerve. 2014. PubMed Abstract \| Publisher Full Text\n\nBossuyt PM, Reitsma JB, Bruns DE, et al.: The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003; 138(1): W1–12. PubMed Abstract \| Publisher Full Text\n\nBossuyt PM, Reitsma JB, Bruns DE, et al.: Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ. 2003; 326(7379): 41–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolloway RG: Improving the flow of diagnostic information. The importance of STARD for authors and readers. Neurology. 2003; 61: 600–1. Publisher Full Text\n\nRossi C, Nattino G, Facchinetti S, et al.: Margherita-PROSAFE project. Promoting patient safety, research, and quality improvement in critical care medicine. Report 2012. Bergamo, Italy: Sestante edizioni; 2013. Reference Source\n\nStevens RD, Dowdy DW, Michaels RK, et al.: Neuromuscular dysfunction acquired in critical illness: a systematic review. Intensive Care Med. 2007; 33(11): 1876–91. PubMed Abstract \| Publisher Full Text\n\nZamperetti N, Latronico N: Clinical research in critically ill patients: the situation in Italy. Intensive Care Med. 2008; 34(7): 1330–2. PubMed Abstract \| Publisher Full Text\n\nBolton CF: Electrophysiologic studies of critically ill patients. Muscle Nerve. 1987; 10(2): 129–35. PubMed Abstract \| Publisher Full Text\n\nBolton CF, Laverty DA, Brown JD, et al.: Critically ill polyneuropathy: electrophysiological studies and differentiation from Guillain-Barre syndrome. J Neurol Neurosurg Psychiatry. 1986; 49(5): 563–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWitt NJ, Zochodne DW, Bolton CF, et al.: Peripheral nerve function in sepsis and multiple organ failure. Chest. 1991; 99(1): 176–84. PubMed Abstract \| Publisher Full Text\n\nDaube JR: AAEM minimonograph #11: Needle examination in clinical electromyography. Muscle Nerve. 1991; 14(8): 685–700. PubMed Abstract \| Publisher Full Text\n\nPetajan JH: AAEM minimonograph #3: motor unit recruitment. Muscle Nerve. 1991; 14(6): 489–502. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Peli E, Botteri M: Critical illness myopathy and neuropathy. Curr Opin Crit Care. 2005; 11(2): 126–32. PubMed Abstract\n\nBuderer NM: Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity. Acad Emerg Med. 1996; 3(9): 895–900. PubMed Abstract \| Publisher Full Text\n\nRich MM, Bird SJ, Raps EC, et al.: Direct muscle stimulation in acute quadriplegic myopathy. Muscle Nerve. 1997; 20(6): 665–73. PubMed Abstract\n\nGoodman BP, Harper CM, Boon AJ: Prolonged compound muscle action potential duration in critical illness myopathy. Muscle Nerve. 2009; 40(6): 1040–2. PubMed Abstract \| Publisher Full Text\n\nvan den Berghe G, Wouters P, Weekers F, et al.: Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001; 345(19): 1359–67. PubMed Abstract \| Publisher Full Text\n\nVan den Berghe G, Wilmer A, Hermans G, et al.: Intensive insulin therapy in the medical ICU. N Engl J Med. 2006; 354(5): 449–61. PubMed Abstract \| Publisher Full Text\n\nMoss M, Yang M, Macht M, et al.: Screening for critical illness polyneuromyopathy with single nerve conduction studies. Intensive Care Med. 2014; 40(5): 683–90. PubMed Abstract \| Publisher Full Text\n\nVan den Berghe G, Schoonheydt K, Becx P, et al.: Insulin therapy protects the central and peripheral nervous system of intensive care patients. Neurology. 2005; 64(8): 1348–53. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Smith M: Introducing simplified electrophysiological test of peripheral nerves and muscles in the ICU: choosing wisely. Intensive Care Med. 2014; 40(5): 746–8. PubMed Abstract \| Publisher Full Text\n\nPuthucheary ZA, Rawal J, McPhail M, et al.: Acute skeletal muscle wasting in critical illness. JAMA. 2013; 310(15): 1591–600. PubMed Abstract \| Publisher Full Text\n\nBatt J, Dos Santos CC, Herridge MS: Muscle injury during critical illness. JAMA. 2013; 310(15): 1569–70. PubMed Abstract \| Publisher Full Text\n\nConnolly BA, Jones GD, Curtis AA, et al.: Clinical predictive value of manual muscle strength testing during critical illness: an observational cohort study. Crit Care. 2013; 17(5): R229. PubMed Abstract \| Publisher Full Text\n\nBanwell BL, Mildner RJ, Hassall AC, et al.: Muscle weakness in critically ill children. Neurology. 2003; 61(12): 1779–82. PubMed Abstract \| Publisher Full Text\n\nStevens RD, Marshall SA, Cornblath DR, et al.: A framework for diagnosing and classifying intensive care unit-acquired weakness. Crit Care Med. 2009; 37(10 Suppl): S299–308. PubMed Abstract \| Publisher Full Text\n\nHough CL, Lieu BK, Caldwell ES: Manual muscle strength testing of critically ill patients: feasibility and interobserver agreement. Crit Care. 2011; 15(1): R43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEikermann M, Latronico N: What is new in prevention of muscle weakness in critically ill patients? Intensive Care Med. 2013; 39(12): 2200–3. PubMed Abstract \| Publisher Full Text\n\nBolton CF: Evidence of neuromuscular dysfunction in the early stages of the systemic inflammatory response syndrome. Intensive Care Med. 2000; 26(9): 1179–80. PubMed Abstract \| Publisher Full Text\n\nZygun DA, Kortbeek JB, Fick GH, et al.: Non-neurologic organ dysfunction in severe traumatic brain injury. Crit Care Med. 2005; 33(3): 654–60. PubMed Abstract \| Publisher Full Text\n\nChristiansen CF, Johansen MB, Christensen S, et al.: Type 2 diabetes and 1-year mortality in intensive care unit patients. Eur J Clin Invest. 2013; 43(3): 238–47. PubMed Abstract \| Publisher Full Text\n\nVinik AI, Nevoret ML, Casellini C, et al.: Diabetic neuropathy. Endocrinol Metab Clin North Am. 2013; 42(4): 747–87. PubMed Abstract \| Publisher Full Text\n\nPerkins BA, Bril V: Diabetic neuropathy: a review emphasizing diagnostic methods. Clin Neurophysiol. 2003; 114(7): 1167–75. PubMed Abstract \| Publisher Full Text\n\nLatronico N, Nattino G, Guarneri B, et al.: Data set for the CRIMYNE-2 study on the validation of peroneal nerve test to diagnose polyneuropathy and myopathy in 121 patients. Figshare. 2014. Data Source" }	[ { "id": "5182", "date": "19 Jun 2014", "name": "Werner J. Z'Graggen", "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\nLatronico and colleagues report the findings of the CRIMYNE-2 study. The preceding CRIMYNE study showed that a simplified electrophysiological investigation (limited to a peroneal nerve conduction study (PENT)) has a sensitivity of 100% and specificity of 67% for the detection of a critical illness polyneuropathy (CIP) and/or critical illness myopathy (CIM) compared to a state of the art electrophysiological investigation. The CRIMYNE-2 study aimed to validate these findings, since the preceding study had some important methodological limitations, which was probably one of the reasons why this simplified approach was not widely accepted by clinicians and researchers. To my opinion, the CRIMYNE-2 study is well designed and achieves its aim. The study was carefully done and confirmed for PENT a sensitivity of 100% and a specificity of 85%. There were no false negative results. The authors conclude that PENT can be used as a screening tool for CIP and CIM and that PENT might be more reliable than other screening methods currently performed, since no patient collaboration is needed. There are some study limitations that should be discussed by the authors in the paragraph about limitations (to some extent the authors mention them in other parts of the manuscript). Based on the current diagnostic criteria for CIM, state of the art electrophysiological investigations allow only the diagnosis of probable CIM, definite diagnosis needs additionally the presence of weakness and histological changes. For CIP, definite diagnosis requires in addition clinical weakness. CRIMYNE-2 did not compare PENT with the criteria for definite CIP/CIM, which could have caused an additional study bias. Furthermore, the authors propose to use PENT for screening without additional repetitive stimulation (which they used in the study). This could lead to an increase of false positive results. Also this aspect should be discussed. The results of CRIMYNE-2 demonstrate that PENT can be used as a screening tool for CIP/M. PENT cannot be used to discriminate between CIP and CIM or concomitant CIP and CIM. The authors should therefore state in their conclusion in the abstract, that PENT can be used as screening test.", "responses": [ { "c_id": "909", "date": "21 Jul 2014", "name": "Nicola Latronico", "role": "Author Response F1000Research Advisory Board Member", "response": "We thank the reviewer for his thoughtful comments. As mentioned in our reply to Dr. Semmler, PENT was conceived as a screening tool, and diagnoses were categorized as normal or abnormal condition with no inference to the differential diagnosis between CIP and CIM. We modified the conclusions in the Abstract to clarify even further this concept, as required by the reviewer." } ] }, { "id": "5212", "date": "23 Jun 2014", "name": "Alexander Semmler", "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, abstract, study design, methods are adequate. Results are sound. Minor comments:The authors compared PENT with standard nerve conduction studies. The diagnostic criteria for CIM/CIP include clinical testing and muscle biopsy. Please discuss why you have not compared your method to the diagnostic standard. The typical patient populations on ICUs are different from the population of the actual study. Patients are often old, and show a high prevalence of diabetes, alcohol abuse and other medical conditions that predispose to polyneuropathy. This may lead to a lot more false positives when only PENT is performed. Please discuss that. Please speculate how PENT could influence the current standard medical practice.", "responses": [ { "c_id": "908", "date": "21 Jul 2014", "name": "Nicola Latronico", "role": "Author Response F1000Research Advisory Board Member", "response": "The authors compared PENT with standard nerve conduction studies. The diagnostic criteria for CIM/CIP include clinical testing and muscle biopsy. Please discuss why you have not compared your method to the diagnostic standard.The CRIMYNE-2 study was a validation study comparing electrophysiological tests, and hence we did not evaluate the muscle strength clinically nor muscle histology, as we declared in the Methods. Possible diagnoses with PENT were normal or abnormal condition with no inference to the differential diagnosis between CIP and CIM. Hence, PENT can only be used as a screening tool, and we further specified this aspect in the Abstract’s conclusion as suggested by the other reviewer. As a high sensitivity test, PENT has a high negative predictive value and is particularly useful in excluding diseased patients if normal. If abnormal, further neurological, electrophysiological and biopsy investigations can be needed to exactly define the nature of the disease. The typical patient populations on ICUs are different from the population of the actual study. Patients are often old, and show a high prevalence of diabetes, alcohol abuse and other medical conditions that predispose to polyneuropathy. This may lead to a lot more false positives when only PENT is performed. Please discuss that.This is an important topic, and we thank to reviewer for giving us the possibility to further discuss it. As we mentioned in the Discussion, the diabetic neuropathy typically involves sensory nerves, motor nerves are involved only in the advanced stage of the disease. Thus, diabetic patients with abnormal PENT would deserve further evaluation. In patients with abnormal PENT, clinical history together with neurological and electrophysiological investigations and appropriate follow-up are essential to identify those patients in whom the CIP complicates a pre-existing diabetic neuropathy. It is worth noting that most of these patients discharged from the ICU are currently left with undefined diagnosis. Please speculate how PENT could influence the current standard medical practice.As a quick, noninvasive and easily repeatable test, PENT can be used as a screening test in all ICU patients with severe critical illness, such as those with protracted ICU stay or mechanical ventilation, those with sepsis and multiple organ dysfunctions or those with clinically identifiable ICU-acquired weakness. This would be a major change in current practice, because, as mentioned above, many critically ill patients do not receive adequate evaluation of acute neuromuscular problems arising during the ICU possibly as a consequence of rapid turnover or other causes 1. As a screening test, PENT may select patients with normal finding who do not require further investigations. Moreover, patients with ICU-acquired weakness have a better outcome if electrophysiology is normal 2. Patients with abnormal PENT will not invariably have CIP or CIM. With further evaluation, they may eventually be diagnosed with peroneal nerve palsy or other peripheral nervous system diseases that merit medical attention. In prospect, electrophysiological alterations of peripheral nerves and muscles may help in instituting timely treatments since they may precede muscle weakness and functional impairment or their diagnosis can be established at an earlier stage 3. Patients in the ICU often are sedated or develop delirium and consciousness disturbance or have acute neurological illnesses precluding the use of clinical tests such as the Medical Research Council score or handgrip dynamometry early in the ICU. References 1. Connolly BA, Jones GD, Curtis AA, Murphy PB, et al.: Clinical predictive value of manual muscle strength testing during critical illness: an observational cohort study.Crit Care.2013; 17 (5): R229 PubMed Abstract \| Free Full Text \| Publisher Full Text 2. Moss M, Yang M, Macht M, Sottile P, et al.: Screening for critical illness polyneuromyopathy with single nerve conduction studies. Intensive Care Med.2014; 40 (5): 683-690 PubMed Abstract \| Publisher Full Text 3. Latronico N, Shehu I, Guarneri B: Use of electrophysiologic testing. Crit Care Med.2009; 37 (10): S316-S320 PubMed Abstract \| Publisher Full Text" } ] } ]	2	https://f1000research.com/articles/3-127
https://f1000research.com/articles/3-98/v1	30 Apr 14	{ "type": "Data Article", "title": "Neurosharing: large-scale data sets (spike, LFP) recorded from the hippocampal-entorhinal system in behaving rats", "authors": [ "Kenji Mizuseki", "Kamran Diba", "Eva Pastalkova", "Jeff Teeters", "Anton Sirota", "György Buzsáki", "Kamran Diba", "Eva Pastalkova", "Jeff Teeters", "Anton Sirota" ], "abstract": "Using silicon-based recording electrodes, we recorded neuronal activity of the dorsal hippocampus and dorsomedial entorhinal cortex from behaving rats. The entorhinal neurons were classified as principal neurons and interneurons based on monosynaptic interactions and wave-shapes. The hippocampal neurons were classified as principal neurons and interneurons based on monosynaptic interactions, wave-shapes and burstiness. The data set contains recordings from 7,736 neurons (6,100 classified as principal neurons, 1,132 as interneurons, and 504 cells that did not clearly fit into either category) obtained during 442 recording sessions from 11 rats (a total of 204.5 hours) while they were engaged in one of eight different behaviours/tasks. Both original and processed data (time stamp of spikes, spike waveforms, result of spike sorting and local field potential) are included, along with metadata of behavioural markers. Community-driven data sharing may offer cross-validation of findings, refinement of interpretations and facilitate discoveries.", "keywords": [ "The hippocampus and entorhinal cortex are essential structures for memory and spatial navigation1–8. Position-tuned cells (‘place cells’) are present in CA1", "CA3 and dentate gyrus regions1", "9. Grid cells", "head direction cells", "and border cells have been described in the dorsomedial entorhinal cortex", "and are critical ingredients of navigation systems5", "7", "8", "10–13. The temporal coordination across the entorhinal cortex and hippocampus is secured by various oscillations", "especially theta", "gamma and sharp wave ripples14–21." ], "content": "Introduction\n\nThe hippocampus and entorhinal cortex are essential structures for memory and spatial navigation1–8. Position-tuned cells (‘place cells’) are present in CA1, CA3 and dentate gyrus regions1,9. Grid cells, head direction cells, and border cells have been described in the dorsomedial entorhinal cortex, and are critical ingredients of navigation systems5,7,8,10–13. The temporal coordination across the entorhinal cortex and hippocampus is secured by various oscillations, especially theta, gamma and sharp wave ripples14–21.\n\nWe recorded activity of neurons in these brain regions while animals performed various tasks, such as linear track, open maze, T-maze with wheel running delay, plus maze and zigzag maze, as well as recordings during sleep in the home cage. Extensive technical descriptions of the data sets described in this document are available in several published papers6,21–27.\n\nSeveral questions related to memory, navigation, spike time patterns, population coding, neuronal interactions, neuronal classification, replay, sleep homeostasis and oscillations have been studied based on this dataset6,21–41. However, this dataset may provide valuable information if subjected to yet further analyses. Improved spike sorting, neuron classification and more sophisticated analyses may extend and refine the initial conclusions and offer insights that were previously missed. For these reasons we provide both unprocessed (wide band) and processed versions of our data. In our experience, all methods have limitations and must undergo continuous revision. We believe that community-driven data sharing, cross-validation of data, unified data formats and large collaborative efforts will facilitate discovery and benefit future progress in neuroscience.\n\n\nMaterial and methods\n\nAll protocols were approved by the Institutional Animal Care and Use Committee of Rutgers University (protocol No. 90-042), and all experiments were performed at Rutgers University. Before surgery, one to four rats were housed in a single home cage (made of plastic; size L = 45 cm, W = 23.5 cm, H = 20 cm). Wood shavings were used as bedding and dry pellets were provided as food. The animals were housed in a temperature controlled (68°F), but not a specific pathogen free, environment under 12:12-hours light:dark cycle where light cycle was from 7AM to 7PM. After surgery, the rats were housed individually, and highly absorbent paper (Techboard, Shepherd Speciality Papers) was used as bedding, and the animal’s health was assessed daily by the experimenters.\n\nDetails of surgery and recovery procedures have been previously described in detail42,43. Eleven Long Evans rats (male, 3–8 months old, 250–400 g) were deeply anesthetized with isoflurane (1–1.5%). In two rats (f01_m and g01_m), two silicon probes were implanted (one in each hemisphere) and targeted CA1 region. In three rats (gor01, pin01 and vvp01), two probes (32- and/or 64-site silicon probes) were implanted in the left dorsal hippocampus, targeted to CA1 and CA3 separately, and advanced over sessions and days through overlying neocortical and hippocampal tissue. The probe positions were: rat pin01: CA3: at a 35 degree angle to coronal plane, centered on 2.8 mm posterior and 2.6 mm lateral to bregma. CA1: 26.5 degree angle to vertical, at a 35 degree angle to coronal, centered on 4.6 mm posterior and 2.4 mm lateral to bregma; rat vvp01: CA3: at a 26.5 degree angle to coronal plane, centered on 2.8 mm posterior and 2.6 mm lateral to bregma. CA1: 26.5 degree angle to vertical, parallel to sagittal plane, centered on 4.4 mm posterior and 2.3 mm lateral to bregma; rat gor01: CA3: at a 26.5 degree angle to coronal plane, centered on 3.1 mm posterior, and 3.0 mm lateral to bregma. CA1: 26.5 degree angle to vertical, at a 45 degree angle to coronal plane, centered on 4.9 mm posterior and 1.5 mm lateral to bregma. In four rats (ec013, ec014, ec016 and i01_m), 32- or 64-site silicon probe(s) were implanted in the right dorsal hippocampus and recorded from CA1, CA3 or dentate gyrus, and another 4-shank silicon probe was implanted in the right dorsocaudal medial entorhinal cortex. In one rat (ec012), one 4-shank silicon probe was implanted in the right dorsocaudal medial entorhinal cortex. In rat ec012, ec013, ec014, and ec016, the probe targeting the entorhinal cortex was positioned such that the different shanks recorded from different layers21 (4.5 mm lateral from the midline; 0.1 mm anterior to the edge of the transverse sinus at a 20–25 degree angle in the sagittal plane with the tip pointing toward the anterior direction). In rat i01_m, the EC probe had 4 shanks and was positioned such that all shanks recorded from the same layer. For the hippocampus probe in rats ec013, ec014 and ec016, the shanks were aligned parallel to the septo-temporal axis of the hippocampus (45 degrees parasagittal), positioned centrally at 3.5 mm posterior from bregma and 2.5 mm lateral from the midline.\n\nFor all silicon probes used, each shank had eight recording sites (160 µm2 each site, 1–3-MΩ impedance), and intershank distance was 200 µm. Recordings sites were staggered to provide a two-dimensional arrangement (20 µm vertical separation)44,45. The individual silicon probes were attached to respective microdrives and moved independently and slowly to the target. Two stainless steel screws inserted above the cerebellum were used as indifferent (reference) and ground electrodes during recordings. At the end of the physiological recordings during the behavioural tasks, a small anodal DC current (2–5 µA, 10 s) was applied to recording sites 1 or 2 days before rats were deeply anesthetized and euthanized by perfusion with 10% formalin solution. The positions of the electrodes were confirmed histologically and reported previously in detail21,24.\n\nAfter the animals recovered from surgery (1 to 2 weeks), physiological signals were recorded during eight different types of behaviours mostly during light cycles (see Table 1).\n\n(1) On an elevated linear track (250 cm × 7 cm), the animals were required to run back and forth to obtain water reward at both ends21. In three animals (gor01, pin01, and vvp01), a similar elevated track was used (170 cm × 6.2 cm, with 22 cm × 22 cm end platforms) that was shortened to 79 or 125 cm in some trials23,24.\n\n(2) In the open field task, the rats chased randomly dispersed drops of water or pieces of Froot Loops (25 mg; Kellogg’s) on an elevated open platform21 (180 cm × 180 cm, 120 cm × 120 cm or 100 cm × 200 cm).\n\n(3) In the rewarded wheel-running task, a wheel (diameter = 29 cm) was attached to a rectangular-shape box (39 cm × 39 cm × 39 cm). The rat was required to run in the wheel continuously for 10 seconds, after which time a piece of Froot Loop was dropped in the box as reinforcement21.\n\n(4) In the alternation task in the T-maze (100 cm × 120 cm) with wheel running delay, the animal was required to run on a wheel attached to the waiting area for 10 sec, after which time the animal had access to the central arm of the T-maze, at the end of which the animal chose to turn right or left. The animal was rewarded with water if the choice was opposite to the previous one6.\n\n(5) In the elevated plus maze (100 cm × 100 cm), the rats were motivated to run to the ends of four corridors, where water was given every 30 s.\n\n(6) In the zigzag maze (100 cm × 200 cm) with 11 corridors, the animals learned to run back and forth between two water wells; 100 µl of water was delivered at each well21,22,25,46.\n\n(7) In the wheel-running in home cage, a wheel (diameter = 29 cm) was attached to a rectangular-shape box (39 cm × 39 cm × 39 cm) which was used as a home cage during the experiment. Rats had free access to the wheel, and ran on the wheel with no reinforcement.\n\n(8) In the sleeping session, the rat slept in the home cage.\n\nFor recording of behaviour (1) to (6), animals were water-scheduled for 23 hours prior to the experiment. Otherwise, both dry food and water were provided ad libitum. For tracking the position of the animals, two small light-emitting diodes, mounted above the headstage, were recorded by a digital video camera (SONY) at 30 Hz resolution.\n\nDetailed information about the recording system and spike sorting has been previously described21,24,42. Briefly, signals were amplified (1,000×), bandpass-filtered (1 Hz–5 kHz) and acquired continuously at 20 kHz (DataMax system; RC Electronics) or 32,552 Hz (NeuraLynx, MT) at 16-bit resolution. After recording, the signals were down-sampled to 1,250 Hz (DataMax system) or 1,252 Hz (NeuraLynx system) for the local field potential (LFP) analysis. In electrophysiological recordings, positive polarity is from zero toward positive values. To maximize the detection of very slowly discharging (‘silent’) neurons47, clustering was performed on concatenated files of several behavioural and sleep sessions recorded at the same electrode position on the same recording day22,25–27. We made extensive use of publicly available analytical and display programs, which were developed in our laboratory (KlustaKwik48 available at http://sourceforge.net/projects/klustakwik/, Neuroscope49 available at http://neuroscope.sourceforge.net/, Klusters49 available at http://klusters.sourceforge.net/, NDmanager49 available at http://ndmanager.sourceforge.net/). The latest available version at the time was used in each case. Spike sorting was performed automatically, using KlustaKwik48, followed by manual adjustment of the clusters, with the help of autocorrelogram, cross-correlogram and spike wave-shape similarity matrix (Klusters software package49). Because none of the existing spike sorting algorithms is completely automated, manual adjustment is necessary48. This inevitably leads to some operator-dependent variability48; therefore, provided clusters are not always identical to those used in our previous publications. Hippocampal principal cells and interneurons were separated based on their burstiness, waveforms and short-term monosynaptic interactions6,17,21,24,42. Classification of principal neurons and interneurons of entorhinal cortical neurons was based on waveforms and short-term monosynaptic interactions, and described previously in detail21. A total of 3,113 (CA1), 882 (CA3), 66 (DG), 491 (EC2), 568 (EC3) and 551 (EC5) principal neurons and 420 (CA1), 198 (CA3), 52 (DG), 85 (EC2), 215 (EC3) and 91 (EC5) interneurons were identified and included in this data set (see Table 2–Table 4).\n\nTop row: animal identifier. Left column: brain region. Brain region EC4 indicates either entorhinal cortex layer 3 or 5 (could not be determined which); region EC? indicates in entorhinal cortex, but without layer assignment.\n\nTop row: animal identifier. Left column: brain region.\n\nTop row: animal identifier. Left column: brain region.\n\nTop row: animal identifier. Left column: behaviour subclass.\n\nThe tip of the probe either moved spontaneously relative to the brain or was moved by the experimenter between recording days to record from potentially different sets of neurons. However, we cannot exclude the possibility that some neurons recorded on different days were identical, because spikes recorded on each day were clustered separately, though in some instances neurons were recorded over multiple days. When we moved the electrodes, we waited for at least an hour before recording in order to stabilize the position of electrodes.\n\nThe data are available50 at CRCNS.org (http://dx.doi.org/10.6080/K09G5JRZ). Details of the data collection, processing and storage of data into files are included with the data set, including scripts useful for processing the data50. Here, we briefly summarize the data description.\n\nThe number of cells recorded from each animal and brain region is shown in Table 2.\n\nMost of the recorded cells were classified as principal neurons or interneurons. The number of cells classified as principal and interneuron are shown in Table 3 and Table 4.\n\nThe 8 types of behaviours (see Behavioural Testing section) were further subdivided into 14 behaviour subclasses based on minor differences (e.g. size of maze) and used as behaviour identifiers in the dataset (Table 1).\n\nThe data were obtained during 442 recording sessions. During each session the animal performed one of the 14 behaviour subclasses. The number of recording sessions and behaviour subclasses used with each animal is shown in Table 5. The description of each behaviour subclass is given in Table 1.\n\nThe data files for each recording session are stored in separate compressed tar archive files (i.e. with extension “tar.gz”). These files are organized into top-level directories, each of which contains data for sessions recorded on the same day using the same animal and electrode placement combination. Data from all sessions recorded from the same animal on the same day were merged for spike sorting. All merged sessions are stored in the same top-level directory in the data set at CRCNS.org. Therefore, the cell identification numbers assigned by the spike sorting are common to all sessions within a top-level directory, and are not specific to individual sessions. Details of the file organization are provided in the document “CRCNS.org hc3 data description” which is included with the data set.\n\nThe metadata describing the data is stored in four tables that are included with the data set. Table cell has information about each spike sorted cell. Table session has information about each experimental session. Table epos contains information about the position of the electrodes. And table file has information about the “.tar.gz” and other files that are in the data set.\n\nThese tables are provided in CSV (comma-separated values) format, Excel format, and as tables in an SQLite database. SQLite (http://www.sqlite.org/) is a free, open source, SQL data base engine available for all common operating systems. These tables are related to each other through a field (named “topdir”), which has the name of top-level directories described above and is common to all four tables. The fields in each of these tables are listed in Listing 1. As described in file “CRCNS.org hc3 data description” the SQLite command interface can be used with these tables to generate summary statistics from the metadata and to locate data files that satisfy particular search criteria (for example, find data for cells of a specific type from a particular brain region and experimental behaviour).\n\nListing 1: Create table statements for tables: cell, session, file and epos. Fields for each of these tables are documented in the comments.\n\n\n\n\nData availability\n\nCRCNS: Multiple single unit recordings from different rat hippocampal and entorhinal regions while the animals were performing multiple behavioral tasks, http://dx.doi.org/10.6080/K09G5JRZ\n\nTerms of data usage: Data on this site is made available only for scientific purposes. Redistribution of the data is not permitted. Any publications derived from the data must cite the data contributors and CRCNS.org as being the source of the data and the original paper(s) that generated the data. Unnecessary downloading of large data files is not permitted. (To minimize demands on the server, only data expected to be useful for your scientific purposes should be downloaded).\n\nPrivacy notice: Occasionally the researchers who contribute data wish to know who has downloaded their data. Upon request we will provide this information to the data contributors. So, if you download data, there is a possibility that your name and email address will be provided to the data contributor. We request that the data contributors only use the information for legitimate scientific purposes (such as determining the frequency of downloads, or contacting users to providing updated information about the data or to explore possible collaborations).", "appendix": "Author contributions\n\n\n\nKM, KD, EP and GB designed the experiments. KM, KD and EP carried out experiments and collected the data. KM collected data from rats ec012, ec013, ec014 and ec016. KD collected data from rats gor01, pin01 and vvp01. EP collected data from rats f01_m, g01_m, i01_m and j01_m. KM carried out all spike sorting and classification of cell types in this dataset. JT prepared documentations for public data release (data sets hc-2 and hc-3) at CRCNS.org. AS prepared an earlier version of documentations for data set hc-2 at CRCNS.org. KM, JT and GB wrote the paper. 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\nThe research was supported by Uehara Memorial Foundation Research Fellowship (KM), Astellas Foundation for Research on Metabolic Disorders Research Fellowship (KM), Japan Society for the Promotion of Science’s Postdoctoral Fellowship for Research Abroad (KM), NIH NS034994 (GB), NIH MH54671 (GB), NIH NS074015 (GB), National Science Foundation Grant S E 0542013 (GB), The Human Frontier Science Program (GB), the James S. McDonnell Foundation (GB), the Kavli Foundation (GB), General Electric, Inc. (GB), HHMI (EP), Patterson Trust (EP), National Science Foundation Grant 0855272 (JT).\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 Gautam Agarwal, Kenneth Harris and members of the Buzsaki lab for support and discussions.\n\n\nReferences\n\nO’Keefe J, Nadel L: The Hippocampus as a Cognitive Map (Oxford: Oxford University Press). 1978. Reference Source\n\nSquire LR: Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. Psychol Rev. 1992; 99(2): 195–231. PubMed Abstract \| Publisher Full Text\n\nEichenbaum H, Dudchenko P, Wood E, et al.: The hippocampus, memory, and place cells: is it spatial memory or a memory space? Neuron. 1999; 23(2): 209–226. PubMed Abstract \| Publisher Full Text\n\nBuzsaki G: Theta rhythm of navigation: link between path integration and landmark navigation, episodic and semantic memory. Hippocampus. 2005; 15(7): 827–840. PubMed Abstract \| Publisher Full Text\n\nMcNaughton BL, Battaglia FP, Jensen O, et al.: Path integration and the neural basis of the ‘cognitive map’. Nat Rev Neurosci. 2006; 7(8): 663–678. PubMed Abstract \| Publisher Full Text\n\nPastalkova E, Itskov V, Amarasingham A, et al.: Internally generated cell assembly sequences in the rat hippocampus. Science. 2008; 321(5894): 1322–1327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuzsaki G, Moser EI: Memory, navigation and theta rhythm in the hippocampal-entorhinal system. Nat Neurosci. 2013; 16(2): 130–138. PubMed Abstract \| Publisher Full Text\n\nMoser EI, Moser MB: Grid cells and neural coding in high-end cortices. Neuron. 2013; 80(3): 765–774. PubMed Abstract \| Publisher Full Text\n\nO‘Keefe J, Dostrovsky J: The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Res. 1971; 34(1): 171–175. PubMed Abstract \| Publisher Full Text\n\nHafting T, Fyhn M, Molden S, et al.: Microstructure of a spatial map in the entorhinal cortex. Nature. 2005; 436(7052): 801–806. PubMed Abstract \| Publisher Full Text\n\nSargolini F, Fyhn M, Hafting T, et al.: Conjunctive representation of position, direction, and velocity in entorhinal cortex. Science. 2006; 312(5774): 758–762. PubMed Abstract \| Publisher Full Text\n\nSolstad T, Boccara CN, Kropff E, et al.: Representation of geometric borders in the entorhinal cortex. Science. 2008; 322(5909): 1865–1868. PubMed Abstract \| Publisher Full Text\n\nSavelli F, Yoganarasimha D, Knierim JJ: Influence of boundary removal on the spatial representations of the medial entorhinal cortex. Hippocampus. 2008; 18(12): 1270–1282. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuzsaki G, Horvath Z, Urioste R, et al.: High-frequency network oscillation in the hippocampus. Science. 1992; 256(5059): 1025–1027. PubMed Abstract \| Publisher Full Text\n\nO‘Keefe J, Recce ML: Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus. 1993; 3(3): 317–330. PubMed Abstract \| Publisher Full Text\n\nBragin A, Jando G, Nadasdy Z, et al.: Gamma (40–100 Hz) oscillation in the hippocampus of the behaving rat. J Neurosci. 1995; 15(1 Pt 1): 47–60. PubMed Abstract\n\nSkaggs WE, McNaughton BL, Wilson MA, et al.: Theta phase precession in hippocampal neuronal populations and the compression of temporal sequences. Hippocampus. 1996; 6(2): 149–172. PubMed Abstract \| Publisher Full Text\n\nBuzsaki G: Theta oscillations in the hippocampus. Neuron. 2002; 33(3): 325–340. PubMed Abstract \| Publisher Full Text\n\nBuzsaki G: Rhythms of the Brain (New York: Oxford University Press). 2006. Reference Source\n\nBuzsaki G, Wang XJ: Mechanisms of gamma oscillations. Annu Rev Neurosci. 2012; 35: 203–225. PubMed Abstract \| Publisher Full Text\n\nMizuseki K, Sirota A, Pastalkova E, et al.: Theta oscillations provide temporal windows for local circuit computation in the entorhinal-hippocampal loop. Neuron. 2009; 64(2): 267–280. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizuseki K, Diba K, Pastalkova E, et al.: Hippocampal CA1 pyramidal cells form functionally distinct sublayers. Nat Neurosci. 2011; 14(9): 1174–1181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDiba K, Buzsaki G: Forward and reverse hippocampal place-cell sequences during ripples. Nat Neurosci. 2007; 10(10): 1241–1242. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDiba K, Buzsaki G: Hippocampal network dynamics constrain the time lag between pyramidal cells across modified environments. J Neurosci. 2008; 28(50): 13448–13456. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizuseki K, Royer S, Diba K, et al.: Activity dynamics and behavioral correlates of CA3 and CA1 hippocampal pyramidal neurons. Hippocampus. 2012; 22(8): 1659–1680. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizuseki K, Buzsaki G: Preconfigured, skewed distribution of firing rates in the hippocampus and entorhinal cortex. Cell Rep. 2013; 4(5): 1010–1021. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizuseki K, Buzsaki G: Theta oscillations decrease spike synchrony in the hippocampus and entorhinal cortex. Philos Trans R Soc Lond B Biol Sci. 2014; 369(1635): 20120530. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBelluscio MA, Mizuseki K, Schmidt R, et al.: Cross-frequency phase-phase coupling between theta and gamma oscillations in the hippocampus. J Neurosci. 2012; 32(2): 423–435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGeisler C, Diba K, Pastalkova E, et al.: Temporal delays among place cells determine the frequency of population theta oscillations in the hippocampus. Proc Natl Acad Sci U S A. 2010; 107(17): 7957–7962. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrosmark AD, Mizuseki K, Pastalkova E, et al.: REM sleep reorganizes hippocampal excitability. Neuron. 2012; 75(6): 1001–1007. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIsomura Y, Sirota A, Ozen S, et al.: Integration and segregation of activity in entorhinal-hippocampal subregions by neocortical slow oscillations. Neuron. 2006; 52(5): 871–882. PubMed Abstract \| Publisher Full Text\n\nItskov V, Pastalkova E, Mizuseki K, et al.: Theta-mediated dynamics of spatial information in hippocampus. J Neurosci. 2008; 28(23): 5959–5964. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nItskov V, Curto C, Pastalkova E, et al.: Cell assembly sequences arising from spike threshold adaptation keep track of time in the hippocampus. J Neurosci. 2011; 31(8): 2828–2834. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKempter R, Leibold C, Buzsaki G, et al.: Quantifying circular-linear associations: hippocampal phase precession. J Neurosci Methods. 2012; 207(1): 113–124. PubMed Abstract \| Publisher Full Text\n\nSchmidt R, Diba K, Leibold C, et al.: Single-trial phase precession in the hippocampus. J Neurosci. 2009; 29(42): 13232–13241. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSullivan D, Csicsvari J, Mizuseki K, et al.: Relationships between hippocampal sharp waves, ripples, and fast gamma oscillation: influence of dentate and entorhinal cortical activity. J Neurosci. 2011; 31(23): 8605–8616. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSullivan D, Mizuseki K, Sorgi A, et al.: Comparison of sleep spindles and theta oscillations in the hippocampus. J Neurosci. 2014; 34(2): 662–674. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchomburg EW, Anastassiou CA, Buzsaki G, et al.: The spiking component of oscillatory extracellular potentials in the rat hippocampus. J Neurosci. 2012; 32(34): 11798–11811. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaxidis J, Mizuseki K, Mason R, et al.: Influence of slow oscillation on hippocampal activity and ripples through cortico-hippocampal synaptic interactions, analyzed by a cortical-CA3-CA1 network model. Front Comput Neurosci. 2013; 7: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStevenson IH, London BM, Oby ER, et al.: Functional connectivity and tuning curves in populations of simultaneously recorded neurons. PLoS Comput Biol. 2012; 8(11): e1002775. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuzsaki G, Mizuseki K: The log-dynamic brain: how skewed distributions affect network operations. Nat Rev Neurosci. 2014; 15(4): 264–78. PubMed Abstract \| Publisher Full Text\n\nCsicsvari J, Hirase H, Czurko A, et al.: Oscillatory coupling of hippocampal pyramidal cells and interneurons in the behaving Rat. J Neurosci. 1999; 19(1): 274–287. PubMed Abstract\n\nVandecasteele M, Royer MS, Belluscio S, et al.: Large-scale recording of neurons by movable silicon probes in behaving rodents. J Vis Exp. 2012; (61): e3568. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCsicsvari J, Henze DA, Jamieson B, et al.: Massively parallel recording of unit and local field potentials with silicon-based electrodes. J Neurophysiol. 2003; 90(2): 1314–1323. PubMed Abstract \| Publisher Full Text\n\nFujisawa S, Amarasingham A, Harrison MT, et al.: Behavior-dependent short-term assembly dynamics in the medial prefrontal cortex. Nat Neurosci. 2008; 11(7): 823–833. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoyer S, Sirota A, Patel J, et al.: Distinct representations and theta dynamics in dorsal and ventral hippocampus. J Neurosci. 2010; 30(5): 1777–1787. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThompson LT, Best PJ: Place cells and silent cells in the hippocampus of freely-behaving rats. J Neurosci. 1989; 9(7): 2382–2390. PubMed Abstract\n\nHarris KD, Henze DA, Csicsvari J, et al.: Accuracy of tetrode spike separation as determined by simultaneous intracellular and extracellular measurements. J Neurophysiol. 2000; 84(1): 401–414. PubMed Abstract\n\nHazan L, Zugaro M, Buzsaki G: Klusters, NeuroScope, NDManager: a free software suite for neurophysiological data processing and visualization. J Neurosci Methods. 2006; 155(2): 207–216. PubMed Abstract \| Publisher Full Text\n\nMizuseki K, Sirota A, Pastalkova E, et al.: Multiple single unit recordings from different rat hippocampal and entorhinal regions while the animals were performing multiple behavioral tasks. CRCNS org. 2013. http://dx.doi.org/10.6080/K09G5JRZ. Data Source" }	[ { "id": "4602", "date": "06 May 2014", "name": "Yoshikazu Isomura", "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 work, Mizuseki et al. provide valuable information on their large-scale data sets of multi-neuronal and local field potential (LFP) recordings from the hippocampus (dentate gyrus, CA3, and CA1) as well as the enthorhinal cortex across the layers (layers 2, 3, 4, 5 etc.) of behaving rats. These data sets have originally been used for a number of their excellent studies for past ten years. Here, they fully summarized the experimental conditions and results in individual recording sessions very clearly for any users to plan to analyze the data. It will be a really useful treasure map for all hippocampus researchers.I have one minor comment. While they showed the number of isolated neurons in each brain area of each animal (Table 2 to Table 4), it is not clear yet how many neurons, at most, were recorded simultaneously. Some people will probably want to analyze spike activity in a population of many neurons. The data would be more available for the higher-order spike analysis if the authors added additional information on the best (or first choice) data set with a sufficient number of well-isolated neurons in the same brain area.", "responses": [ { "c_id": "903", "date": "15 Jul 2014", "name": "Kenji Mizuseki", "role": "Reader Comment", "response": "Indeed one should be able to know the number of neurons recorded in each recording sessions before downloading the data. Following the reviewer’s advice, we added tables giving the number of spikes recorded from each neuron during each session to the metadata tables describing the data set and we provide examples in document “crcns-hc3-data-description” of how to use these to find sessions useful for the researcher’s purpose before downloading the data. In addition, we include a new set of documents (in file “crcns-hc3-cell-counts.zip”) that lists the sessions with the most cells from each region (for both all cells and well isolated units) and also lists - for all sessions - the number of cells recorded from each region (both all cells and well isolated). These documents can also be used to select sessions with the largest number of simultaneously recorded neurons from particular regions before downloading the data." } ] }, { "id": "4607", "date": "06 May 2014", "name": "Shuzo Sakata", "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 important report for the neuroscience community regarding huge data sets recorded from the dorsal hippocampus and dorsomedial entorhinal cortex of behaving rats. Because the authors have already published a number of papers with the data sets, I have no doubt that this paper and their shared data will contribute to the further advancement of this field. This manuscript is clearly written in details. This article can be improved by clarifying the following minor points in the section Data collection and cell-type classification:1. Cluster qualityIt is not clear whether the quality of single units was assessed objectively, for example, by measuring isolation distance.2. Cell-type classificationBecause now researchers including the authors can classify cell-types optogenetically, it would be better to comment on potential pitfalls of the classification procedure used here. For example, it is difficult to exclude the possibility that some of “principal” neurons can be interneurons.", "responses": [ { "c_id": "901", "date": "15 Jul 2014", "name": "Kenji Mizuseki", "role": "Reader Comment", "response": "\"It is not clear whether the quality of single units was assessed objectively, for example, by measuring isolation distance.\"To check the quality of spike sorting, we calculated isolation distance1, an interspike interval index \"R2/10”2 and a fraction of interspike intervals less than 2 msec3. In the revised manuscript, we explain that these values are available for each neuron so that one can choose well-isolated neurons by using these isolation quality measurements before downloading the data. We also mention that an initial quality control filtering was done in selecting cells to include in the data set: 7943 cells were detected, but only 7736 of them were included in the data set; 207 were not included because they were judged to have insufficient quality. “Because now researchers including the authors can classify cell-types optogenetically, it would be better to comment on potential pitfalls of the classification procedure used here. For example, it is difficult to exclude the possibility that some of “principal” neurons can be interneurons.”This is an important point. Even though our classification of principal neurons and interneurons is supported by short-term spike cross-correlogram and consistent with previous reports4 5 6 misclassification of pyramidal cells and interneurons is inevitable to some extent by our method. For example, some types of interneurons with wide wave shapes (e.g. somatostatin positive ineterneurons) might have been classified as pyramidal cells by our method7. The cell-type classification method should be verified and refined by optogenetical tools in the future.7 8 9 10 11 We discuss this point in the revised manuscript. References 1. Harris KD, Hirase H, Leinekugel X, Henze DA, et al.: Temporal interaction between single spikes and complex spike bursts in hippocampal pyramidal cells.Neuron.2001; 32 (1): 141-149 PubMed Abstract \| Publisher Full Text 2. Fee MS, Mitra PP, Kleinfeld D: Automatic sorting of multiple unit neuronal signals in the presence of anisotropic and non-Gaussian variability.J Neurosci Methods.1996; 69 (2): 175-188 PubMed Abstract \| Publisher Full Text 3. Takehara-Nishiuchi K, McNaughton BL: Spontaneous changes of neocortical code for associative memory during consolidation.Science.2008; 322 (5903): 960-963 PubMed Abstract \| Publisher Full Text 4. Barthó P, Hiras H, Monconduit L, Zugaro M, et al.: Characterization of neocortical principal cells and interneurons by network interactions and extracellular features.J Neurophysiol.2004; 92 (1): 600-608 PubMed Abstract \| Publisher Full Text 5. Quilichini P, Sirota A, Buzsáki G.: Intrinsic circuit organization and theta-gamma oscillation dynamics in the entorhinal cortex of the rat.J Neurosci.2010; 30 (33): 11128-11142 PubMed Abstract \| Free Full Text \| Publisher Full Text 6. Sirota A, Montgomery S, Fujisawa S, Isomura Y, et al.: Entrainment of neocortical neurons and gamma oscillations by the hippocampal theta rhythm.Neuron.2008; 60 (4): 683-697 PubMed Abstract \| Free Full Text \| Publisher Full Text 7. Royer S, Zemelman BV, Losonczy A, Kim J, et al.: Control of timing, rate and bursts of hippocampal place cells by dendritic and somatic inhibition.Nat Neurosci.2012; 15 (5): 769-775 PubMed Abstract \| Publisher Full Text 8. Madisen L, Mao T, Koch H, Zhuo JM, et al.: A toolbox of Cre-dependent optogenetic transgenic mice for light-induced activation and silencing.Nat Neurosci.2012; 15 (5): 793-802 PubMed Abstract \| Free Full Text \| Publisher Full Text 9. Stark E, Koos T, Buzsáki G.: Diode probes for spatiotemporal optical control of multiple neurons in freely moving animals.J Neurophysiol.2012; 108 (1): 349-363 PubMed Abstract \| Free Full Text \| Publisher Full Text 10. Stark E, Eichler R, Roux L, Fujisawa S, et al.: Inhibition-induced theta resonance in cortical circuits.Neuron.2013; 80 (5): 1263-1276 PubMed Abstract \| Free Full Text \| Publisher Full Text 11. Roux L, Stark E, Sjulson L, Buzsáki G.: In vivo optogenetic identification and manipulation of GABAergic interneuron subtypes.Curr Opin Neurobiol.2014; 26: 88-95 PubMed Abstract \| Free Full Text \| Publisher Full Text" } ] }, { "id": "4604", "date": "06 May 2014", "name": "Andrew P. Maurer", "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\nMizuseki and colleagues provide a description of 442 datasets (more than 200 hours) of hippocampal in vivo recordings. These datasets provide cell classification as well as the raw data, in case users wish to return to the high-sample traces and re-cluster the data on their own. The results of different analyses from the database have been published, although all analysis possibilities have not been exhausted. While this is an atypical review to write (as any suggestions on improvement to the database seem to be akin to looking a gift horse in the mouth), I am hoping the authors can discuss the ramifications of providing such an extensive database. While any database has limitations (e.g. sometimes determining the EEG units is a bit of an exercise with these data in the present format), it is perhaps more advantageous and constructive to discuss what the community can do to make the most use of it. As the authors state “Community-driven data sharing may offer cross-validation of findings, refinement of interpretations and facilitate discoveries” and the challenges are now directed to those who are interested in participating in future analyses. In hopes that this will set-forth a new era of data-sharing, I hope that the authors can discuss their open database in a manner that parallels Giorgio Ascoli’s discussion on sharing neural reconstruction files (Ascoli, 2006). It should be noted that I do not expect the authors to have comprehensive answers to each of my comments below, but it might be beneficial if they would provide some initial thoughts to seed further discussion. Some points that the authors may wish to discuss include:One barrier to sharing data is “the fear of being scooped” (Ascoli, 2006). For example, scientific progress will be dramatically increased through parallel (and hopefully, collaborative) data analysis. Are the authors concerned about being “scooped”? What about a group of researchers unknowingly using the database to conduct an analysis that is also a student’s PhD project? This hypothetical student may still be acquiring the expertise to keep pace with more seasoned researchers. “An often unspoken resistance to the sharing… data is born out of concern for criticisms and mistakes” (Ascoli, 2006). This is an unprecedented event that Mizuseki and colleagues have set forth, by providing a comprehensive catalog of data in an unabashed manner. The first point I want to touch upon is the level of raw exposure in releasing data. I have never met a neuroscientist who believes that anyone else could’ve conducted their analyses better than themselves (save a modest few), but to find those who believe that the authors “didn’t look for the right thing” could make up a small battalion. Mizuseki and colleagues invite critics to their doorstep. Perhaps this is more similar to posing nude as a model for artist. By placing data online, it comes with judgment and the potential to be proved wrong. What do the authors believe the convention is if others follow in their footsteps? For example, if group A shares their database after publishing their results, group B downloads and analyzes their data in new light and finds the opposite results, what is the appropriate manner in handling the situation? The self-correction aspect of science is also accelerated when data is openly-shared. It remains to be seen how situations like this should/will be handled. There is also a chance for the data to be used in order to stifle or impede publication when the result is dubious. That is, should it be considered “fair-play” for a reviewer to use the same database and conduct a similar analysis with results that contradict a submitted manuscript’s results? A tactic such as this only seems appropriate in “open-review formats”. “A final barrier to sharing digital reconstructions relates to the reluctance to lose or give a competitive edge” (Ascoli, 2006). The release of this immense database will surely be the stronghold of many new assistant professors who are still in the initial stages of setting up a physiology laboratory. Moreover, I can see these data being used in laboratories that are heavily analysis driven and limited in their own capacity for high-density in vivo recordings. This increase in the number of people analyzing data invites competitors for the authors as well as their neutral peers. Sheer logic dictates that the authors are not afraid of competition (otherwise, why share the data? Please note that I am a cynic and have been taught by many reviewers that “scientific altruism” is as abundant as snarks and unicorns). Why should researchers not be afraid of others using this database to compete? Is it believed that these data will be used for collaboration? What can be done to emphasize collaboration across laboratories when using the database? How is authorship handled when multiple groups use these data? If I spend my time analyzing these data, only to have it published under a “group project name” (similar to Sir William Timothy Gowers’ Polymath project), do I put it on my CV? The paper explicitly states that large downloads are prohibited. Does this mean that I should not download all the data? Is it OK to use these data for a class project? If so, is it more appropriate for the professor to disseminate it to the students or should the students make their own CRCNS account? Finally, more of an afterthought but along similar lines: should there be a collaborative processing code library that should be developed and maintained in parallel with the use of these data (similar to GitHub or SourceForge)? I absolutely do not expect the authors to have complete answers to these questions nor should they carry the sole responsibility of determining the general conventions of what constitutes use versus misuse, but I do think that it is worth hearing their general thoughts. As this is the largest and most comprehensive database of in vivo hippocampal and entorhinal physiology to become available to the general scientific community, there will be an immense ramification. Scientific replication/external validation is an immediate and positive application of this database. As I have cited Timothy Gowers above, I think it would be best to leave off with his opinion of open data-sharing and collaboration: “It feels as though this process is to normal research as driving is to pushing a car.” (http://gowers.wordpress.com/2009/02/01/questions-of-procedure/).As a field we have the opportunity to compete or collaborate. I hope that these data facilitate cross-laboratory collaboration where two groups are reticent to share their own data. For those that are interested in embracing the collaborative spirit, the CRCNS website also has a “marketplace” section where ideas and potential collaborations can be discussed. Finally, I applaud the authors for this unprecedented act of scientific altruism. I hope this will be a platform that accelerates our understanding of the entorhinal-hippocampal circuitry through collaboration.", "responses": [ { "c_id": "900", "date": "15 Jul 2014", "name": "Kenji Mizuseki", "role": "Reader Comment", "response": "One barrier to sharing data is “the fear of being scooped” (Ascoli, 2006). For example, scientific progress will be dramatically increased through parallel (and hopefully, collaborative) data analysis. Are the authors concerned about being “scooped”? What about a group of researchers unknowingly using the database to conduct an analysis that is also a student’s PhD project? This hypothetical student may still be acquiring the expertise to keep pace with more seasoned researchers.We absolutely agree that “the fear of being scooped” is a big concern and for this reason some people feel strong resistance against data sharing. On the other hand, this fear is not specific to data sharing: similar experiments and analysis are underway by many groups independently every day with or without knowing that that is also a student or postdoc’s project in other groups. Our tentative feeling is that, if the research group (principal investigator and his/her students and postdocs) is happy to share data publicly, we do not see any reason not to do so. If the group is not willing to share data due to the fear of being scooped, we do not see any reason to force them to share the data. “An often unspoken resistance to the sharing… data is born out of concern for criticisms and mistakes” (Ascoli, 2006). This is an unprecedented event that Mizuseki and colleagues have set forth, by providing a comprehensive catalog of data in an unabashed manner. The first point I want to touch upon is the level of raw exposure in releasing data. I have never met a neuroscientist who believes that anyone else could’ve conducted their analyses better than themselves (save a modest few), but to find those who believe that the authors “didn’t look for the right thing” could make up a small battalion. Mizuseki and colleagues invite critics to their doorstep. Perhaps this is more similar to posing nude as a model for artist. By placing data online, it comes with judgment and the potential to be proved wrong. What do the authors believe the convention is if others follow in their footsteps? For example, if group A shares their database after publishing their results, group B downloads and analyzes their data in new light and finds the opposite results, what is the appropriate manner in handling the situation? The self-correction aspect of science is also accelerated when data is openly-shared. It remains to be seen how situations like this should/will be handled. We fully agree that releasing data may be judged as “risky” by many. We have added this information to the text and cite Ascoli (2006). However, we think that this ‘self-correction’ process is a sound one and accelerates the progress in science in the long run. Since the scientific community has only a limited history of data sharing, good manner of data sharing should be formed in the community through experience. There is also a chance for the data to be used in order to stifle or impede publication when the result is dubious. That is, should it be considered “fair-play” for a reviewer to use the same database and conduct a similar analysis with results that contradict a submitted manuscript’s results? A tactic such as this only seems appropriate in “open-review formats”.Of course this should be judged case by case, but in general we think the situation the reviewer described is a fair-play and even a sound one. Tactics and politics should not be part of scientists’ vocabulary. We need to assume that all of our colleagues are honest and they assume the same about us. “A final barrier to sharing digital reconstructions relates to the reluctance to lose or give a competitive edge” (Ascoli, 2006). The release of this immense database will surely be the stronghold of many new assistant professors who are still in the initial stages of setting up a physiology laboratory. Moreover, I can see these data being used in laboratories that are heavily analysis driven and limited in their own capacity for high-density in vivo recordings. This increase in the number of people analyzing data invites competitors for the authors as well as their neutral peers. Sheer logic dictates that the authors are not afraid of competition (otherwise, why share the data? Please note that I am a cynic and have been taught by many reviewers that “scientific altruism” is as abundant as snarks and unicorns). Why should researchers not be afraid of others using this database to compete? Is it believed that these data will be used for collaboration? What can be done to emphasize collaboration across laboratories when using the database? How is authorship handled when multiple groups use these data? If I spend my time analyzing these data, only to have it published under a “group project name” (similar to Sir William Timothy Gowers’ Polymath project), do I put it on my CV?We do not have the answers to all these questions, of course. We can talk only about our opinions. Sharing data has two compelling reasons (1) The experiments were largely performed using public money, therefore it is only natural to return our results (data) to public at some point. (2) We cannot exhaust the every single possible analysis by ourselves, and other researchers may have much better idea and capability to analyze the data set to provide new insights. The paper explicitly states that large downloads are prohibited. Does this mean that I should not download all the data? Is it OK to use these data for a class project? If so, is it more appropriate for the professor to disseminate it to the students or should the students make their own CRCNS account? Finally, more of an afterthought but along similar lines: should there be a collaborative processing code library that should be developed and maintained in parallel with the use of these data (similar to GitHub or SourceForge)? CRCNS terms allow downloading all the data. The guidelines (not preconditions) suggest that \"Unnecessary downloading of large data files is not permitted.\" CRCNS just wants to avoid that industry and for-profit organizations will inappropriately use public data. If someone needs to download all the files for their scientific purpose, then it's OK. We modified the terms to allow sharing of downloaded data amongst direct colleagues working with the data or to students in a classroom. The CRCNS.org website does have forum that allows users to upload attachments that contain code or to reference code being developed at other sites, such as GitHub. Thanks for the suggestion to have a collaboratively developed code library that is more directly associated with data sets." } ] }, { "id": "4603", "date": "07 May 2014", "name": "James Knierim", "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 data set will be a valuable resource for investigators who wish to test hypotheses about hippocampal function and interaction with entorhinal cortex at the level of single unit and LFP physiology. I have not investigated the data base carefully to ensure its utility, as I assume the authors have done so. My comments are limited to a few questions about their F1000Research article describing the data base.'Animal surgery' - second paragraph:Do the authors mean \"in the coronal plane?\" It is not clear whether the tetrodes were angled medially or laterally in that plane, or whether the authors mean that the tetrodes were angled anteriorly or posteriorly to the coronal plane. Please clarify here and in other locations in the text how the tetrodes were angled. 'Animal surgery' - final paragraph:Does the data base contain histological figures? If not, are they easily identifiable and accessible from published reports? It would be very useful to ensure that the precise location of tetrodes could be made available to investigators. Table 1, 'elevated linear track':Shouldn't the fixed angles be 74 degrees (i.e. +/- 37 degrees)? 'Behavioural testing' item (5):Were the rats motivated to run the corridors in any particular order? Was there any working memory component to the trials?", "responses": [ { "c_id": "898", "date": "14 Jul 2014", "name": "Kenji Mizuseki", "role": "Reader Comment", "response": "\"Do the authors mean \"in the coronal plane?\" It is not clear whether the tetrodes were angled medially or laterally in that plane, or whether the authors mean that the tetrodes were angled anteriorly or posteriorly to the coronal plane. Please clarify here and in other locations in the text how the tetrodes were angled.\"Indeed the description of probe implantation in the previous manuscript was not detailed enough. What we meant by ‘to the coronal plane’ was ‘a plane formed by silicone probe shanks was 35 degrees to the coronal plane, roughly along the septo-temoporal axis’. We provide accurate and complete descriptions of probe implantation in the revised manuscript. \"Does the data base contain histological figures? If not, are they easily identifiable and accessible from published reports? It would be very useful to ensure that the precise location of tetrodes could be made available to investigators.\" The data set does not contain histological figures, but all available histological figures are reported in the previous publications. In the revised manuscript, we described the relationship between the animal identifiers in this paper and the animal identifiers in the previous publications so that one can easily identify the histology of the animal. \"Shouldn't the fixed angles be 74 degrees (i.e. +/- 37 degrees)?\"Yes, indeed this should be 74 degrees and we fixed it. \"Were the rats motivated to run the corridors in any particular order? Was there any working memory component to the trials?\"The plus maze had two open and two closed arms, and each arm had a water cup at the outer end of it. Before the session started, all the four water cups were filled with water (~20μl). Once the animal consumed water in all of the water cups, water was given in two water cups in closed or open arms depending on which type of arms the animal visited and consumed the water lastly, that is, if the animal consumed the water in the closed arm lastly water was provided in water cups in the open arms (and vice versa). Afterword, when the animal consumed the water in the two water cups, water was given in the two water cups in the opposite type of arms to encourage the animal to visit all the wells with similar probability. We added this information in the revised manuscript." } ] } ]	1	https://f1000research.com/articles/3-98
https://f1000research.com/articles/2-93/v1	22 Mar 13	{ "type": "Research Article", "title": "Improved performance of sequence search algorithms in remote homology detection", "authors": [ "Adwait Govind Joshi", "Upadhyayula Surya Raghavender", "Ramanathan Sowdhamini", "Adwait Govind Joshi", "Upadhyayula Surya Raghavender" ], "abstract": "The protein sequence space is vast and diverse, spanning across different families. Biologically meaningful relationships exist between proteins at superfamily level. However, it is highly challenging to establish convincing relationships at the superfamily level by means of simple sequence searches. It is necessary to design a rigorous sequence search strategy to establish remote homology relationships and achieve high coverage. We have used iterative profile-based methods, along with constraints of sequence motifs, to specify search directions. We address the importance of multiple start points (queries) to achieve high coverage at protein superfamily level. We have devised strategies to employ a structural regime to search sequence space with good specificity and sensitivity. We employ two well-known sequence search methods, PSI-BLAST and PHI-BLAST, with multiple queries and multiple patterns to enhance homologue identification at the structural superfamily level. The study suggests that multiple queries improve sensitivity, while a pattern-constrained iterative sequence search becomes stringent at the initial stages, thereby driving the search in a specific direction and also achieves high coverage. This data mining approach has been applied to the entire structural superfamily database.", "keywords": [ "distant relationships", "function annotation", "sequence search algorithms", "protein superfamilies" ], "content": "Introduction\n\nProtein sequence databases have grown enormously in recent times. Understanding protein homology within such huge sets of sequences requires tracing the divergence by mutation, substitution, insertion and deletion of residues1,2. Homologous proteins reflect similarity at sequence and structural levels, implying functional similarity3. This level of similarity broadens into the superfamily and the ways to deduce such relationships differ for both protein sequence and structure information4,5. There are different databases that organize sets of homologous proteins or protein superfamilies based on protein sequence and structure. These databases primarily employ protein domain information present in a sequence or structure. SCOP is a database that organizes the protein structural domain data in different hierarchical levels based on structural and functional information6. Many sequence search strategies use SCOP domains as a starting point for homology detection, focusing mainly at the superfamily level7,8. Structure-based classification is helpful to explore sequence space and helps in functional assignments by association of protein sequences9.\n\nOf the several methods developed for protein homology detection, the popular BLAST10 algorithm uses heuristics to search sequences and is able to detect close homologues, but fails in a few instances to establish relationships between distantly related proteins. To detect remote homologues, several methods such as, PSI-BLAST10 based on profiles, Hidden Markov Model (HMM)-based methods like HMMSEARCH and Jackhmmer11,12, pattern-based methods like PHI-BLAST13, intermediate sequence search methods such as Cascade PSI-BLAST14 and phylogenetic tree based searches like Treesearch15 have been developed. Methods such as CHASE incorporate some of the above methods in a combined manner to reinforce the sequence search16. Each of the above methods can be optimized for better performance by customizing their parameters and the way they are implemented for sequence searches. For instance, PSI-BLAST is an iterative PSSM (Position Specific Scoring Matrix)-based remote homology detection method. In Cascade PSI-BLAST the search is iterated for several generations in a cascaded manner to improve remote homology detection14. It is important to select appropriate start points for sequence searches, especially for protein families and superfamilies, as different start points can result in different coverage. Park and co-workers have shown that remote homology detection is enhanced threefold for a set of related sequences in the form of a profile than merely searching with a single sequence as a query17. Anand and co-workers emphasized the use of multiple PSSMs as better detectors of remote homologues compared to a single query18. Thus, a search strategy can be designed to improve remote homology detection by choosing appropriate method(s) and starting point(s) for the search and by further optimizing the parameters.\n\nWe have considered multi-member superfamilies from the PASS2 database19. PASS2 is a database of structural alignments of protein domains in a SCOP superfamily which share less than 40% mutual sequence identity. The strategy lies in using all the members of PASS2 superfamily, i.e., multiple members, and searching against the NR-Db (Non Redundant Database) available at the NCBI protein resource. We present our analysis based on a multiple query approach (MQ) for PSI-BLAST and PHI-BLAST. While using PHI-BLAST, for each query we obtain a set of patterns to initiate multiple searches per query, thereby adding an extra dimension for multiple patterns. Search approaches are evaluated at different coverage levels and a comparison of the two approaches using the two methods is presented. The methods are then applied to the entire PASS2 database.\n\n\nMaterials and methodology\n\nDifferent structural classes of proteins described in SCOP were considered for remote homology detection. PASS219 based on the SCOP database and the ASTRAL compendium20, uses protein structural entries from the SCOP superfamily with less than 40% mutual sequence identity. Three superfamilies each from four structural classes from the PASS2 database were selected for this study (Table 1). Each superfamily contains a varied number of members, with some containing a single family and others with more than one family. The classification of families is based on SCOP hierarchies; PASS2 does not consider this classification. Therefore, different members within a PASS2 superfamily may be listed in different families in SCOP. The sequence search strategy devised, described below, was implemented for data mining the sequence homologues for structural superfamilies. Structural members from all the 1961 PASS2 superfamilies were used to search for sequence homologues.\n\nEach member of the PASS2 superfamily was selected as a query for homologue detection and used to search against the NR-Db available at the NCBI ftp site (ftp://ftp.ncbi.nih.gov/blast/db/FASTA/nr.gz). Two popular methods, PSI-BLAST and PHI-BLAST, from the BLAST 2.2.23+ package, were used for sequence searches (ftp://ftp.ncbi.nih.gov/blast/executables/blast+/2.2.23/ncbi-blast-2.2.23+-x64-linux.tar.gz) (see Figure 1 for a flow-chart).\n\nPSI-BLAST was used for each query until convergence or a maximum of 20 iterations. The expectation value threshold (E-value: parameter –e) and inclusion threshold (parameter –h) were both set to 0.001. In all the iterations, all the alignment regions from each of the hit sequences were considered for further validation steps. Following this, a set of all unique sequence homologues (hits with a unique GenInfo Identifier (GI)) was recorded for each member. Sets of hits from all the members in a superfamily were pooled and filtered to generate a cumulative set of unique GIs for each superfamily. Employing all members for homology detection for a superfamily is essentially the multiple query (MQ) approach in the sequence search strategy. For all the PASS2 superfamilies, data mining was carried out using this MQ approach and PSI-BLAST on the latest NR-Db version (June, 2012 release) to populate a sequence homologue database for structural superfamilies7. PHI-BLAST requires a query and a pattern pair as an input. A set of patterns was identified for each query and each query-pattern pair was used for the PHI-BLAST search. The parameters and their values were maintained as in PSI-BLAST searches. The use of multiple patterns for all the multiple queries (MPMQ approach), for superfamily level homology detection was tested on the 12 PASS2 superfamilies.\n\nHomologues were searched for query sequence patterns, using stringent PSI-BLAST criteria (maximum 10 iterations, 10-10 E-value), in NR-Db. A set of hits with 60–90% sequence identity was selected for the pattern generation process. For queries with <10 hits, the sequence identity window was relaxed to 40–95%. These hits were aligned using ClustalW 2.021. An in-house program, MOTIFS (the program can be made available upon request), was used to identify motifs from the Multiple Sequence Alignment (MSA). The program accepts a multiple sequence alignment of protein sequences as input and employs a DAYHOFF-type amino acid exchange matrix to score amino acid replacements for every pair of sequences. The average pairwise amino acid exchange scores are recorded at every alignment position. In general, alignment positions with high scores means a highly conserved position. A motif is reported for a stretch of three or more consecutive positions with complete conservation or favorable substitution (based on an identical or similar residue in a given position in the MSA and the average score). For all such identified motifs, the stretch of amino acids from all the sequences is reported with their length and position corresponding to the first sequence in the alignment. The motif generated is screened for the following conditions:\n\nLength of the motif should be ≥ 3\n\nAt each position, there should be only ≤3 different amino acid preferences (reported by MOTIFS); if not, then that position is denoted as ‘X’, where ‘X’ refers to any amino acid at that position\n\nWhen there is an occurrence of two consecutive ‘X’s (variable residues), the pattern is divided into two separate patterns\n\nNo pattern starts or ends with ‘X’\n\nAny pattern can have a maximum length of 15 amino acid positions, since above this length there is little increment in the number of hits obtained\n\nAfter the motif passes through all the above filters, it is converted into a PROSITE format pattern, and used for PHI-BLAST along with its query22.\n\nThe PASS2 database contains structure-based sequence alignments which are used to build a HMM for each superfamily. Since these HMMs are built using structural alignments, they are highly sensitive in validating the sequence homologue obtained for a superfamily member as a true positive (TP). A library of all PASS2 superfamily HMMs was consulted using HMMSCAN, from the HMMER 3.0 package (ftp://selab.janelia.org/pub/software/hmmer3/3.0/hmmer-3.0-linux-intel-x86_64.tar.gz), to validate all the sequence homologues through either of the sequence approaches (MQ and MPMQ) discussed above11. This approach of validation had earlier been effective in 80% of the examples23.\n\nIn the process of data mining for all PASS2 superfamilies using the MQ approach, additional validation was used for the homologues which failed to associate with the correct superfamily HMM. A single query HMM was built for all PASS2 members which were grouped together to generate a library. Thus, a two-step validation was carried for all PASS2 superfamilies.\n\nCoverage for both of the sequence search approaches was based on different levels. Firstly, a cumulative set of validated hits (Cum-TP) obtained for each superfamily should contain all the PASS2 members comprising that particular superfamily. Secondly, hits corresponding to the sequences of all the SCOP members for a given superfamily should be part of the Cum-TP set. Since, PASS2 contains superfamilies where no two members within a superfamily had >40% sequence identity, the coverage at the SCOP level reflects the sensitivity of the search strategy. Any cross-superfamily hits detected were considered as false positives (FP), even if they belong to the same fold. Thirdly, since sequences of all structural entries present in Protein Data Bank (PDB) obtained as hits are not classified into SCOP, all structural entries were inspected through HMM validation whether they associate to a single PASS2 superfamily HMM with no cross-superfamily associations. Finally, all the protein sequence homologues devoid of the above mentioned levels, with no structural information, were validated by HMMSCAN.\n\nIn each superfamily, a set of cumulative true positives (Cum-TP) was formed with hits having unique GI obtained from different members. Therefore, from the set of all hits obtained for a superfamily (All Hits), the number of validated hits, i.e. Cum-TP, was used to calculate positive prediction value (PPV) for all 12 superfamilies in both approaches. The PPV is calculated as follows:\n\n\n\nA member with highest number of TPs was identified as a best representing sequence (BRS) and its TP count was also recorded (BRS-TP) for all superfamilies. Corresponding to Cum-TP, a ratio of BRS-TP/All Hits was also calculated. Using this value and Cum-TP, a percentage gain in the coverage (PGC) was recorded:\n\n\n\n\nResults\n\nThe sequence search strategy devised for remote homology detection was tested and implemented.\n\nSelection of parameters. 12 structural superfamilies were considered for testing the sequence search strategy. These superfamilies span different classes (α, β, α/β, α+β) of proteins, as per the SCOP definitions. In the MQ approach, all the members from each of the 12 selected superfamilies were used as inputs for PSI-BLAST to search against the NR-Db. The performance of PSI-BLAST was optimized for different parameters. The E-value (parameter –e) and inclusion threshold (parameter –h) were optimized to 0.001, after testing it on a range of 1 to 10-10. An optimized E-value ensures better coverage with fewer FP. The number of iterations (parameter –j) was set to 20, after testing for values 5, 10 and 20. Some queries may converge within 5 iterations where the superfamily was less diverse as in the case of 50203 (bacterial enterotoxin). However, for certain superfamilies like 55239 (RuBisCo-small subunit), none of the members could converge searches within 20 iterations explaining the abundance of sequence homologues for such proteins in the NR-Db. The rest of the parameters were set to default.\n\nQuery retention. In the profile-based iterative (PSI-BLAST) run, the sensitivity and specificity depends upon the quality of the PSSM generated per iteration. If the PSSM gets corrupted, then the PSI-BLAST search drifts to FPs by inclusion of non-homologous sequences. This can be traced by retention of the query in the PSSM throughout the iterations of a PSI-BLAST run. A query retained until the end reflects good optimization of search parameters and little corruption of the PSSM. To study the query retention for each member in a superfamily, the presence of query was inspected in all iterations for the PSI-BLAST run. The number of iterations was divided into four bins as 25%, 50%, 75% and 100% of total iterations for which the PSI-BLAST run lasted or converged. If a PSI-BLAST run converged at the 16th iteration, upper limits for these four bins were 4th, 8th, 12th and 16th iterations, respectively. The query was inspected for the occurrence in any of the four bins for all members of a superfamily. The frequency of queries last observed in any of the four bins is plotted in Figure 2(a). It reflects that out of a total of 70 queries from all 12 superfamilies, 45 queries were retained until the last iteration and 23 queries were not retained beyond 50% of iterations. However, it was observed that close homologues of the member query (sequence identity >90%) were retained until the last bin, thereby driving the PSI-BLAST search in the correct direction. Query drift was observed in only two out of the total 70 members from 12 different superfamilies.\n\n(A) Frequency of query retention for 12 superfamilies and (B) for all the queries from all the superfamilies in PASS2 database. PASS2 database is classified as SMS (Single-Member Superfamily), TMS (Two-Member Superfamily) and MMS (Multi-Member Superfamily) on the basis of number of members in each superfamily.\n\nIn the data mining of all structural superfamilies within the PASS2 database using the MQ approach, a similar trend was observed for the entire set of 10569 queries arising from 1961 superfamilies. 65% of queries were observed in the fourth bin, thereby ensuring the parameters selected for this scale-up of a MQ approach were acceptable. At times, the loss of query in the PSSM could still be beneficial, if the resultant PSSM is enriched by neighboring families, thus improving the coverage at the superfamily level. The PASS2 database includes the classification of superfamilies as single-member (SMS), two-member (TMS) and multi-member (MMS) based on the number of members in each superfamily. A distribution for the query retention in the four bins for these sections is shown in Figure 2(b). 68% of MMS queries have >50% query retention. However, in the MQ approach, few of the members in MMS are retained until the fourth bin to ensure better sequence search at the superfamily level.\n\nEach member and pattern pair was used for sequence searches using PHI-BLAST. Similar to PSI-BLAST, the E-value (parameter –e) and inclusion threshold (parameter –h) were optimized to 0.001. The maximum number of iterations allowed was set to 20. While generating the patterns, it was ensured that patterns are specific and stringent to query. The length of the patterns generated varies from 3 to 40 amino acid positions. Longer lengths reflect the long conserved stretches in the homologues selected for pattern generation. Multiple patterns for each query and multiple queries per superfamily (MPMQ approach) were used in the sequence searches. This gives rise to multiple start points to search in sequence space for each superfamily. Although PHI-BLAST essentially follows an iterative protocol like PSI-BLAST, the hits identified in the first iteration differ as they are constrained to have the pattern specified along with the query. This ensures that the hits will be used to create the first PSSM for the rest of the iterative search process.\n\nThe sequence hits obtained through MQ and MPMQ approaches were validated for TP and then inspected for the coverage at different levels for all 12 superfamilies.\n\nCoverage for PASS2 members. Every member was examined for its coverage of all the PASS2 members in the sequence search. All the members achieved full coverage when searched in the cumulative set of hits derived from sequence searches of all the PASS2 members. Every member of the two superfamilies, RuBisCo small subunit (SCOP code: 55239) and mannose-binding lectins (SCOP code: 51101), could identify all other members. But for the rest of the superfamilies, every member was unable to identify all the PASS2 members. Although, single members in such instances were unable to cover all the PASS2 members cumulatively, presence of all members gave rise to 100% coverage, thereby stating the importance of the use of multiple queries (MQ).\n\nCoverage for SCOP superfamily members. The coverage for all the SCOP superfamily members was inspected. The ratio of the number of sequence homologues of SCOP superfamily entries identified as TP to the total number of SCOP superfamily members is plotted as shown in Figure 3 (blue bars). Almost all of the superfamilies obtained >90% coverage for all SCOP superfamily members. The only exception was the bacterial enterotoxins superfamily with two families (SCOP code: 50203), where coverage was 25%. In the PASS2 database, this superfamily contained nine members from one SCOP family, and could identify most of its SCOP family members. But none of the members could identify cross-family members. However, there was only one member from the other family which was not sufficient to identify all the family members. This lack of coverage can be overcome if more members are included from the second family. For such a diverse superfamily, it is apparent that unless we use an MQ approach, it is difficult to achieve a reasonable coverage. In the rest of the superfamilies, some members were able to identify cross-family members while some could not identify any cross-family member. In either case, all of the SCOP family members were covered. Cumulatively, all PASS2 members are strong enough to cover most the SCOP superfamily members underlining the high sensitivity of MQ approach.\n\nThe blue bars correspond to the percentage of SCOP members identified as hits. The red bars correspond to percentage ratio of TP (True Positives) identified by BRS (Best Representative Sequence) and All Hits (BRS-TP/All Hits). The green bars correspond to percentage ratio of Cumulative TP identified for a superfamily and All Hits (Cum-TP/All Hits).\n\nCoverage for the structural (PDB) entries. The sequences of structural entries present in PDB were part of the All Hits set. These are newer entries which are yet to be accounted for in SCOP. After PASS2-HMM validation, they were inspected for which superfamily they belong to and if they could associate to the query member’s superfamily. There were no cross-superfamily connections, highlighting the specificity of the sequence search through the MQ approach.\n\nCoverage with respect to the TP sequence homologues. While following the MQ approach over a huge database like NR-Db, it is necessary to achieve a maximum number of homologues (TP) and minimum unrelated sequences (FP). The use of multiple queries ensures a high TP count, while appropriate stringency of the E-value restricts occurrence of FPs. Following validation based on PASS2 HMMs, a set of Cum-TP was obtained and a ratio of Cum-TP/All Hits was calculated. Similarly, a BRS-TP/All Hits ratio was also calculated. Both these values are expressed as percentages as shown in Figure 3 (green and red bars respectively). The percentage of SCOP entries identified is also represented in Figure 3 for each superfamily. Considering the Cum-TP/All Hits ratio, Figure 3 shows the extent of coverage achieved by the BRS. It is clear that none of the BRS could achieve more than an 80% TP share of all the TPs identified at superfamily level. The average coverage for BRS across all superfamilies was ~50%. However, the Cum-TP/All Hits ratio was >90% for most of the superfamilies. In most of the cases, the false positives were hits which could not get associated with any PASS2-HMM, but did not associate with any other superfamily apart from query superfamily. This may partly be due to the fact that the superfamily was so diverse that the PASS2-HMM itself was getting noisy due to the diversity or else the hit obtained was too distant to get validated using the HMM validation cut-offs. On average, false positives were <3%. The results strongly suggest that the MQ approach to sequence data mining helps to increase coverage at the complete superfamily level. For superfamilies which have multiple families (47336, 55031), none of the BRS could obtain a BRS-TP/All Hits ratio >35%, but cumulative coverage by all queries for a complete superfamily is excellent (100% Cum-TP/All Hits ratio). Even in the case of a single-family superfamily (55031 - bacterial exopeptidase dimerisation domain superfamily), all the members have attained a 100% Cum-TP/All Hits ratio.\n\nThe coverage of the MPMQ approach for all 12 superfamilies was equal to or better than the MQ approach at each level discussed above. All the PASS2 members were identified cumulatively by all the members. The coverage increased for all the superfamilies with respect to the SCOP superfamily (except for the 50203 superfamily which performed poorly even in the MQ approach). The sequences of structural (PDB) entries identified were associated with the query superfamily PASS2-HMM. Similar to PSI-BLAST, PHI-BLAST results also revealed that the use of multiple queries are beneficial for achieving better coverage at PASS2, SCOP and PDB level.\n\nThe MPMQ approach was contrasted with the MQ approach (Table 2). The high PPV value suggests that both approaches were able to discern between the TP and FP. The MPMQ approach has relatively more start points than the MQ approach, employing pattern-constrained PHI-BLAST. The BRS identified in PSI-BLAST was also the same in the PHI-BLAST results for most of the superfamilies. The performance of both approaches was compared by plotting percentage coverage achieved using the BRS-TP/All Hits and Cum-TP/All Hits (Figure 4). The Cum-TP/All Hits was highest in all superfamilies for the PHI-BLAST runs. This was chosen as 100% coverage (% saturation), to compare performance of BRS in MQ and MPMQ and Cum-TP/All Hits for the MQ approach. The Cum-TP/All Hits (MPMQ) was almost double for 55307 and 51971. This difference between MPMQ and MQ approaches was >20% for 9 out of 12 superfamilies. Although the MPMQ approach has promising results, compared with the MQ approach, it is computationally very expensive. Each member in a superfamily generates multiple patterns, hence the higher the number of query-pattern pairs means the number of PHI-BLAST runs to be performed for the given superfamily increases proportionally.\n\nThe blue and green bars correspond to TP obtained for BRS (BRS-TP) and cumulatively (Cum-TP) using all queries in the superfamily respectively using PSI-BLAST. The red and purple bars correspond similarly to TP obtained for BRS (BRS-TP) and cumulatively (Cum-TP) for all queries in the superfamily respectively using PHI-BLAST. All values are %saturation considering Cum-TP for PHI-BLAST (highest number of TP) as the saturation value for each superfamily.\n\nThe performance of BRS in both approaches was also inspected. BRS for MPMQ achieved close to 80% coverage in 9 out of 12 superfamilies, with an exceptionally low value of 25% for the 51971 superfamily (a multi-family superfamily). Considering multi-family superfamilies (47336, 50203, 51679, 51971), the BRS of MPMQ could not cross 50% coverage compared with the Cum-TP/All Hits of MPMQ. This reinforces and lends support to the strategy of using multiple queries for sequence searches at the superfamily level. The MPMQ BRS was able to identify more TPs than Cum-TP obtained through MQ for the 51069, 55239, 51351 superfamilies, which are single-family superfamilies. This implies that BRS in isolation, through PHI-BLAST, can drive sequence searches attaining better coverage in such superfamilies, achieving a trade-off between coverage and computational time. However, performance of BRS is highly superfamily-specific.\n\nGenDiS (Genomic Distribution of protein structural domain Superfamilies) is our previous database of sequence homologues based on PASS2, where sequence homologues for structural members are organized with respect to their genomic distribution at the superfamily level7. In an attempt to populate GenDiS, the MQ approach was implemented to search in sequence databases starting from the entire 1961 PASS2 superfamilies. PASS2 classification of SMS, TMS and MMS relies on the number of constituent members in a superfamily. For each superfamily a percentage gain in coverage (PGC) value due to improvised strategies, in comparison with simple PSI-BLAST runs, was calculated based on the preliminary results on 731 MMS (see Methods E). The PGC was divided into four quarters and the frequency of superfamilies belonging to each quarter is shown in Figure 5a. It was found that 532 superfamilies obtain >50% PGC. This observation strongly suggests the necessity of the MQ approach for remote homology detection, where the BRS alone collects fewer homologues. The MQ approach has to be implemented, especially for huge databases like NR-Db, to obtain improved coverage at the superfamily level. An inspection of the relation between number of members and PGC for MMS is shown in Figure 5b, revealing an increase in PGC value with a rise in the number of members. The PGC values are extremely high for diverse and large superfamilies with more than 30 members.\n\n(A) Frequency of superfamilies based on PGC (Percentage Gain in Coverage) value divided into 4 bins (0–25%, 26–50%, 51–75%, 76–100%). (B) The PGC value plotted against number of members for all superfamilies.\n\n\nDiscussion\n\nWe have employed structural entries from the PASS2 database to understand and improve sequence data mining techniques, where the searches are specifically meant for distant homologues19. SCOP is a good benchmark dataset to evaluate many homology detection methods6, but removing redundancy with respect to similar entries reduces the search time. However, with the use of the MQ approach for searching sequence space there is substantial improvement in coverage, albeit with an increase in search time. Use of PASS2 superfamily members as query is a trade-off, between use of all the SCOP members as query (ideal for best coverage) and a single query for the entire structural superfamily (ideal for saving computational time).\n\nMQ grants an additional advantage of coverage, which is revealed by considering coverage at different levels. A multi-member SCOP superfamily like 48345, with multiple families, and where each query gives an exclusive set of hits through both the methods, use of multiple queries is inevitable to achieve a better or even complete coverage. These findings comply with previous findings based on different search methods17,24.\n\nTo date, many methods and approaches have been standardized for sequence searches at different levels of database complexity8,16–18,24. PSI-BLAST is in many instances a primary choice for homology detection. Efforts by different groups have enhanced the usage of PSI-BLAST, improving its sensitivity in remote homology detection14,25. PHI-BLAST, employing an initial pattern–hit initiation, enhances the search method before continuing over profile-based searches in subsequent iterations24. In the present study this has been shown to increase the coverage by >20% for many of the selected superfamilies. The simultaneous use of different methods greatly improves coverage, and the combinations of other methods needs to be investigated.\n\nThe choice of BRS identified from MQ may provide a better option to run PHI-BLAST for a single query. This is beneficial considering an increase in coverage, but it is necessary to acknowledge that, multiple queries and their multiple patterns increase computational time tremendously for a given superfamily. Although for a multi-member, multi-family superfamily, this may not effectively work, this approach can serve as a good trade-off between coverage and computational time for a superfamily with a single family. Also, for a very large superfamily (>20 members), the BRS-PHI-BLAST will be a good option to choose, however the choice of any of the discussed approaches for sequence search becomes superfamily specific.\n\n\nConclusion\n\nIn the post-genomic era, there is a strong need to devise sequence search strategies for effective functional annotation. Functional annotation transfer and convincing establishment of sequence relationships are bottlenecks at low sequence identities. Hence, there is a constant quest for a robust computational program which is greedy on coverage and stringent in eliminating false positives. Remote homology detection through sequence searches has been addressed by various methods and approaches. Considering a protein structural superfamily, when an attempt to view its spread in the sequence space is made, one can resort to various methods and approaches. This study shows that a MQ approach proves beneficial for increasing the coverage. The PASS2 database, which accounts for SCOP superfamily members that are structurally aligned, becomes a good start point to scan sequence space for good coverage at the complete SCOP superfamily level. Instead of using all SCOP members, using only PASS2 members with a MQ gives rise to good coverage in remote homology detection. In this paper, we have further compared the MQ approach using PSI-BLAST to a MPMQ using PHI-BLAST. The MPMQ approach totally outperforms the MQ approach. Use of patterns makes the initial pattern-hit initiation stringent, thus elevating the specificity of the search24. The study also indicates that a best representative sequence, which performs better with the PHI-BLAST than the MQ PSI-BLAST, can help save computational time and achieve better coverage. But this is superfamily-specific and for multi-family superfamilies, a query from one family may not be sufficient to associate with hits obtained from a query of a different family within a given superfamily. It is, therefore, necessary to resort to a MQ or MPMQ approach for getting reasonable coverage. The PGC obtained from data mining for all multi-member superfamilies from PASS2 points to the necessity of using multiple queries. Thus, different methods and different approaches can be used to improve the sequence searches during remote homology detection.", "appendix": "Author contributions\n\n\n\nAGJ was involved in the benchmarking of sequence searches for 12 superfamilies. Both AGJ and USR contributed the sequence searches for the entire superfamily databases. RS was involved in the design and critical analysis of the work. AGJ wrote the first draft of the manuscript and both USR and RS contributed to the improvement of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAGJ is supported by a fellowship (09/860(0104)/2010-EMR-I) from the Council of Scientific and Industrial Research, India. USR is supported by Open Source Drug Discovery (OSDD) grant (OSDD/HCP001/11FYP/2011–12/124) from the Council of Scientific and Industrial Research (CSIR), India. RS thanks CSIR for OSDD Grant and Department of Biotechnology, India for financial support in the form of Centre for Excellence Grant (BT/01/COE/09/01).\n\n\nAcknowledgments\n\nAll authors thank NCBS for infrastructural and general support.\n\n\nReferences\n\nThorne JL: Models of protein sequence evolution and their applications. Curr Opin Genet Dev. 2000; 10(6): 602–605. PubMed Abstract \| Publisher Full Text\n\nOrengo CA, Thornton JM: Protein families and their evolution-a structural perspective. Annu Rev Biochem. 2005; 74: 867–900. PubMed Abstract \| Publisher Full Text\n\nLee D, Redfern O, Orengo C: Predicting protein function from sequence and structure. Nat Rev Mol Cell Biol. 2007; 8(12): 995–1005. PubMed Abstract \| Publisher Full Text\n\nWhisstock JC, Lesk AM: Prediction of protein function from protein sequence and structure. Q Rev Biophys. 2003; 36(3): 307–340. PubMed Abstract \| Publisher Full Text\n\nWatson JD, Laskowski RA, Thornton JM: Predicting protein function from sequence and structural data. Curr Opin Struct Biol. 2005; 15(3): 275–284. PubMed Abstract \| Publisher Full Text\n\nMurzin AG, Brenner SE, Hubbard T, et al.: SCOP: a structural classification of proteins database for the investigation of sequences and structures. J Mol Biol. 1995; 247(4): 536–540. PubMed Abstract \| Publisher Full Text\n\nPugalenthi G, Bhaduri A, Sowdhamini R: GenDiS: Genomic Distribution of protein structural domain Superfamilies. Nucleic Acids Res. 2005; 33(Database issue): D252–D255. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWistrand M, Sonnhammer E: Improved profile HMM performance by assessment of critical algorithmic features in SAM and HMMER. BMC Bioinformatics. 2005; 6: 99. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDessailly BH, Redfern OC, Cuff A, et al.: Exploiting structural classifications for function prediction: towards a domain grammar for protein function. Curr Opin Struct Biol. 2009; 19(3): 349–356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltschul SF, Madden TL, Schäffer AA, et al.: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25(17): 3389–3402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEddy SR: A new generation of homology search tools based on probabilistic inference. Genome Inform. 2009; 23(1): 205–211. PubMed Abstract \| Publisher Full Text\n\nJohnson LS, Eddy SR, Portugaly E: Hidden Markov model speed heuristic and iterative HMM search procedure. BMC Bioinformatics. 2010; 11: 431. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang Z, Miller W, Schäffer AA, et al.: Protein sequence similarity searches using patterns as seeds. Nucleic Acids Res. 1998; 26(17): 3986–3990. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSandhya S, Chakrabarti S, Abhinandan KR, et al.: Assessment of a rigorous transitive profile based search method to detect remotely similar proteins. J Biomol Struct Dyn. 2005; 23(3): 283–298. PubMed Abstract \| Publisher Full Text\n\nRehmsmeier M, Vingron M: Phylogenetic information improves homology detection. Proteins. 2001; 45(4): 360–371. PubMed Abstract \| Publisher Full Text\n\nAlam I, Dress A, Rehmsmeier M, et al.: Comparative homology agreement search: an effective combination of homology-search methods. Proc Natl Acad Sci U S A. 2004; 101(38): 13814–13819. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark J, Karplus K, Barrett C, et al.: Sequence comparisons using multiple sequences detect three times as many remote homologues as pairwise methods. J Mol Biol. 1998; 284(4): 1201–1210. PubMed Abstract \| Publisher Full Text\n\nAnand B, Gowri VS, Srinivasan N: Use of multiple profiles corresponding to a sequence alignment enables effective detection of remote homologues. Bioinformatics. 2005; 21(12): 2821–2826. PubMed Abstract \| Publisher Full Text\n\nBhaduri A, Pugalenthi G, Sowdhamini R: PASS2: an automated database of protein alignments organised as structural superfamilies. BMC Bioinformatics. 2004; 5: 35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChandonia JM, Hon G, Walker NS, et al.: The ASTRAL Compendium in 2004. Nucleic Acids Res. 2004; 32(Database issue): 189D–192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLarkin MA, Blackshields G, Brown NP, et al.: Clustal W and Clustal X version 2.0. Bioinformatics. 2007; 23(21): 2947–2948. PubMed Abstract \| Publisher Full Text\n\nBairoch A: PROSITE: a dictionary of sites and patterns in proteins. Nucleic Acids Res. 1991; 19(Suppl) 2241–2245. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMutt E, Mitra A, Sowdhamini R: Search for Protein Sequence Homologues that Display Considerable Domain Length Variations. Int J Knowl Dis Bioinform. 2011; 2(2): 55–77. Publisher Full Text\n\nBhaduri A, Ravishankar R, Sowdhamini R: Conserved spatially interacting motifs of protein superfamilies: application to fold recognition and function annotation of genome data. Proteins. 2004; 54(4) 657–670. PubMed Abstract \| Publisher Full Text\n\nLee MM, Chan MK, Bundschuh R: Simple is beautiful: a straightforward approach to improve the delineation of true and false positives in PSI-BLAST searches. Bioinformatics. 2008; 24(11): 1339–1343. PubMed Abstract \| Publisher Full Text" }	[ { "id": "3732", "date": "11 Mar 2014", "name": "Saikat Chakrabarti", "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\nConcept and strategy of the work:In this study the authors have chosen 12 superfamilies from PASS2 database and collected the unique sets of hits generated by PSI-BLAST and PHI-BLAST against the non-redundant protein database (NR-Db) and checked how many of the sequences from the same superfamily are returned. There are quite a number of already published reports that dealt with the same problem but used different dataset and search algorithms. The same research group has already published a report where they have worked improvement of PSI-BLAST using different parameters1,2. Though neither the concept nor the strategy is novel, the search for a reliable method for detection of remote homologues is crucial. The report provides good information about the performance of PSI-BLAST and PHI-BLAST with a defined set of parameters and also compares their performance. This could be useful information for future researchers who want to use the same homology search algorithm. I believe clarification of the following points will improve the article. Major Comments:The strategies (MQ) and (MPMQ) are not selected by choice. It is an obvious process. If one wants to check the remote homologues in a superfamily level; one has to take all the sequences of the superfamily. Also MP is an obvious requirement for PHI-BLAST. So MPMQ and MQ are rather obvious requirement for running the programs effectively. But it has been reported other way round. The authors have used a small fraction of the total dataset (12 out of 1961 PASS2 super-families) for demonstration of their algorithm. Testing the algorithm on a larger dataset will be helpful for establishing the efficacy of the algorithm. As mentioned in the manuscript, the result for 1961 families should be produced. Also the basis for choosing only those 12 superfamilies is not clearly mentioned. Would the result be similar for any random 12 superfamilies? Also, will this process return the same robustness if we use the same superfamilies from SCOP, where the number of sequences is greater? The authors have demonstrated the efficacy of the two methods by means of bar graphs (figures 3 and 4), which represent the coverage percentage of each of the approaches of MQ and MPMQ. It would be more informative if the authors provided receiver operating characteristics (ROC) plots to show the rate of detection true positives (TPs) and false positives (FPs) by the methods. In reference to figure 4, the authors should explain a little more about why they have considered the coverage percentage of MPMQ approach as the saturation percentage. In this way the analysis is biased towards PHI-BLAST. In my opinion, the representation of the true coverage in percentage for both methods would be more informative. It is commendable how the authors have been able to identify all possible patterns that might be a property of a superfamily. However, my suggestion would be to use other third party motif-finding programs (apart from MOTIFS for which no published references were provided) and consider the consensus motif predicted by all such software for a particular super-family. Information about those motifs will be helpful. Care should be taken to ensure that the motifs are biologically and statistically relevant. In the Validation section, HMM profiles for all PASS2 superfamilies were made and homologues were scanned against the HMM. The filtering criteria here are not mentioned. They also mention that they have used two-step validations. But actually the second step is used for the hits which fail to associate with any superfamily by level 1 search. So the validation is not two-step but one-step for each set. The title states “Improved performance of sequence search algorithms in remote homology detection” but there is no improvement measure taken for enhancing the performance. And both the algorithms are well established and widely used for remote homology search. So what is the new improvement of the algorithms made by the authors? On page 6 maximum length of motif is mentioned as 40 and in the Materials and methodology section (page 4) the same is mentioned as 15. Which one has actually been used? In figure 5b, where increase in PGC is compared with number of member in families, the relation between the two is not clear. There are few families with more than 30 members. But only those few families have very high PGC. So a statistical relation is required. From the graph it looks like the algorithm is biased towards multi-member families which contain a high number (more than 30) of members. The efficiency of the program must be tested on single member, two-member and those multimember families which have less than 30 members.Minor Comments:-The authors could throw some light on the fact as to why they have not used all the family members of a SCOP superfamily as has been mentioned in PASS2. For example, in table 1 the number of families mentioned by the authors for SCOP superfamily 47336 is 7, whereas PASS2 lists at 8. It is not really clear what the coverage of the Best Representative Sequence (BRS) of a superfamily dataset says about the performance of the method. In the collection of the hit pool, unique sets of homologues (hits) for each superfamily were taken. That means each protein in the sets of hits is associated with one query protein of the superfamily protein only. How is that done? The method is not clear. In graph 2A and 2B the y-axes are comparable so they should be in same scale. Both in figure 3 and 4 if the one-member, two-member and multi-member families were marked separately the result would be more prominent. In the explanation of coverage, where the differences of the superfamilies are explained in terms of their constituent families, actual examples or a case study may be more helpful.", "responses": [ { "c_id": "856", "date": "11 Jun 2014", "name": "Ramanathan Sowdhamini", "role": "Author Response", "response": "We thank Dr. Saikat Chakrabarti for their useful and encouraging comments. The MQ or MPMQ approach, whilst being obvious choices, are also computationally demanding and this analysis on 12 superfamilies was a conscious attempt to measure the extent of gain/loss of coverage. We have now introduced a Venn diagram that provides a cumulative measure of coverage of these different approaches (Figure 4-A of the revised manuscript). We do find similar trends when applied on the 1961 superfamilies. We had meant this as an approach, not an algorithm, and we have now modified the title as well. Our responses to specific comments:Major Comment 3: We agree that statistical measures such as ROC curves are often employed to compare methods and approaches. However, since our sequence searches are against the non-redundant sequence database, despite our independent validations, there is no clear negative dataset nor is there is a ‘gold standard’. So it is not possible to measure FPR and compare them for different approaches. Major Comment 4: A comparison of hits by different approaches shows that MPMQ achieves the highest coverage. We now provide a Venn diagram (Figure 4-A of the revised manuscript) that provides a quantitative idea of the performance of different approaches in terms of accumulation of validated hits. Major Comment 6: It is true that the validation is a one-step procedure since single-query HMM validation was invariably performed for validation of all the hits; however, it was necessary only when full-Sf HMM failed to validate. Major Comment 7: We are sorry that the title has been misleading. We had meant to say ‘sequence search approaches’ since there is indeed no new algorithm involved. We have now modified the title. However, the improvement in performance is to arrive at a good balance between sequence coverage and computational time. Our analysis shows that it is possible to obtain good coverage by employing multiple queries followed by multiple patterns for the best representative query. Minor query 1:We have used all the members from PASS2 for all the sequence searches. In case of superfamily 47336, the current updated version of PASS2 has 8 members whereas the earlier (2008) version has been used for the analysis of 12 superfamilies. Therefore there is a difference in the number of members. Minor query 4: We are sorry about this and we have now modified Figure 2 in the revised manuscript. Minor query 6: Thanks for this suggestion. We are now including one more example in the revised manuscript to convey practical biological application." } ] }, { "id": "3320", "date": "17 Apr 2014", "name": "Mallur Srivatsan Madhusudhan", "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 study the authors have sought to improve the alignment protocols that make use of standard tools such as PSI-BLAST and PHI-BLAST to find protein sequences that are remotely related to a query sequence. These distant relationships would be useful in annotating protein functions and designing experiments.The authors of this study use multiple queries (MQ) as a starting point to detect protein homologous relationships. The results from MQ are compared against another search that makes use of multiple patterns deduced from multiple queries (MPMQ). The protocol has been benchmarked on a set of 12 superfamilies consisting of different fold types. The manuscript is clearly presented and the plots and tables make their points clearly and unambiguously.I wish that the authors would clarify the following points:For any given query sequence what would be the best strategy to adopt: MQ or MPMQ? How would a potential user make this judgement? The prescriptions (conclusions section) from the text are a little ambivalent, especially to a non-expert user/reader. Demonstrating the efficacy of the search protocol over only 3 examples of each type of fold (alpha, beta, alpha/beta and alpha+beta) may not be sufficient. The authors should look to increase the benchmark size or show that that the current benchmark is statistically representative. While the authors have not discussed this in depth, the implication appears to be that the search results appear to be independent of fold type. Could the authors comment on whether this is true for other types of SCOP fold families (trans-membrane proteins etc)? In the two parts of Figure 2, it is apparent that most PSSMs diverge (non-retention of query), if at all, within the first 50% of the iterations. Only very seldom does this happen afterwards. Is there a lesson one can learn from this result? Could future search strategies make use of this observation? Admittedly not all divergent PSSMs are detrimental to the finding of homologues, but is there a way one can control for this during the search? It would be instructive to have a somewhat more quantitative description of the process in the abstract. I think that reader would benefit in knowing how much better the two protocols are (‘X’ more sequence homologues identified; or TP rate goes up by ‘X’ and FP rate reduces by ‘Y’ - or something along these lines).The same quantifiable measures should also be mentioned in the concluding section.Minor pointsIn the pattern generation sub-section of the Methods section, the authors say that in the event of inadequate number of hits they relax the stringent criteria of allowing hits between 60-90% in sequence identity to include sequences in the range 40-95%. Why is the range 40-95% not used uniformly? Is it the case that the pattern recognition program suffers from a drastic decrease in efficiency when decreasing sequence identity from 60 to 40%? What happens in cases where patterns are >15 amino acids in length? How does the program deal with this? Does it break the patterns into subsets of lengths that have a maximum length of 15?", "responses": [ { "c_id": "855", "date": "11 Jun 2014", "name": "Ramanathan Sowdhamini", "role": "Author Response", "response": "We thank Dr. Madhusudhan for providing constructive comments and interests. We had applied these enhanced sequence search strategies to 1961 superfamilies and we find similar trends. We have now mentioned recommended strategies and parameters in the Conclusion section. Having said that, some of the strategies, search directions and parameters depend on the sequence dispersion and population of families and few decisions are highly family-specific.We have the following responses to specific comments: Major query 1: The MQ approach is easiest to follow as the user needs only the input sequences for a given superfamily. If the superfamily is very diverse with many members (>10), the MQ approach can be adopted, perhaps for representative sequences, to achieve a trade-off between computational time and coverage. We have now mentioned this in Conclusions of the revised manuscript.Minor query 1: It is difficult to obtain patterns qualifying stringent cut-offs followed (please refer to Methods section) when the sequence identity is reduced to 40% that will be applicable for all the superfamilies. It was important to ensure that patterns do reflect conservation, thereby rendering specificity to the PHI-BLAST searches. However, in cases where there were few homologues to deduce a conserved pattern even at 60% sequence identity, 40% sequence identity cut-off had to be employed. Minor query 2:If a pattern is >15 amino acids in length, the long pattern is broken into subsets of patterns which qualify all the stringent cut-offs and have a maximum length of 15." } ] } ]	1	https://f1000research.com/articles/2-93
https://f1000research.com/articles/3-57/v1	14 Feb 14	{ "type": "Research Article", "title": "Dosing of ceftriaxone and outcomes after spontaneous bacterial peritonitis", "authors": [ "Laura Mazer", "Elliot B. Tapper", "Gail Piatkowski", "Michelle Lai", "Laura Mazer", "Gail Piatkowski", "Michelle Lai" ], "abstract": "Background: Spontaneous bacterial peritonitis (SBP) is a common, often fatal affliction for cirrhotic patients. Despite all clinical trials of ceftriaxone for SBP using 2g daily, it is often given at 1g daily.Aim: We evaluated outcomes of SBP as a function of ceftriaxone dosage.Methods: A retrospective cohort of all patients who received ceftriaxone for SBP (greater than 250 neutrophils in the ascites).Results: As opposed to 1 gram, median survival is longer for patients receiving 2 grams (228 days vs. 102 days (p = 0.26) and one year survival is significantly higher (p = 0.0034). After adjusting for baseline Model for End Stage Liver Disease (MELD) score, however, this difference was no longer significant. Similarly, there was a significantly shorter length of intensive care for patients receiving 2 g (0.59 ± 1.78 days vs. 3.26 ± 6.9, p = 0.034), odds ratio 0.11 (95% CI 0.02 - 0.65). This difference, too, was no longer significant after controlling for the MELD score - odds ratio 0.21 (95% CI 0.04 - 1.07). Additionally, 70% of patients received at least one additional antibiotic; over 25 different medications were used in various combinations.Conclusions: We recommend fastidious antibiotic stewardship for patients with cirrhosis and that efforts should be made to standardize the treatment of SBP. The complexity of antibiotic regimens to which cirrhotic patients are exposed must be studied further and rationalized. Patients receiving 2 g of ceftriaxone may require fewer intensive care days and enjoy an improved 1 year survival compared to those receiving 1 g daily.", "keywords": [ "Cirrhosis", "Quality Improvement", "Liver Disease", "Ascites" ], "content": "Introduction\n\nAscites is the most common hepatic decompensation, occurring in 50% of cirrhotic patients followed for over a decade1. The development of ascites heralds a vulnerable time of sharply increased mortality for patients with liver disease – related in large part to spontaneous bacterial peritonitis (SBP)2,3. SBP is an infection of the ascitic fluid that occurs in 10–30% of patients with ascites4. Fatal in as many as 32.6% of cases, SBP can have a profound effect on the tenuous hemodynamics of patients with cirrhosis5. Exacerbating the arterial underfilling that results from the splanchnic vasodilation of cirrhosis, SBP may lead to a decrease in cardiac output such that it can no longer satisfy the needs of a kidney that is already vasoconstricted6. The result is the hepatorenal syndrome which is often devastating. SBP with renal injury is fatal in 42% of patients7.\n\nSBP is caused by translocation of gastrointestinal organisms into the ascitic fluid, most commonly Escherichia coli, Klebsiella pneumoniae and Streptococcus pneumoniae. As such, third generation cephalosporins are amongst the best studied antibiotics in this setting, with ceftriaxone as the drug of choice where cefotaxime is not available. Studied as a treatment for SBP in clinical trials for 25 years, the doses employed were either 1 g every 12 hours or 2 g every 24 hours given intravenously for 5 to 10 day courses8–14.\n\nAt our center, we have found that ceftriaxone is often given at 1 g daily either in reference to online resources from other major teaching institutions or because 1 g is the general preset dose for this antibiotic as generated by the electronic ordering system15. (http://clinicalpharmacy.ucsf.edu/idmp/adult_guide/empiric_guide/intraabd_hosp_frame.htm, last accessed 1-12-2014). The outcomes of SBP as a function of ceftriaxone dosage – 1 g daily versus 2 g daily – have never been evaluated. It is unknown what effect the dosage of ceftriaxone has on the control of SBP or on mortality. Neither the American Association for the Study of Liver Disease (AASLD) nor the European Association for the Study of the Liver (EASL) guidelines on SBP management explicitly comment on the dosing of ceftriaxone for this indication16,17.\n\nHerein we present the results of a retrospective review of the outcomes of SBP stratified by dose of ceftriaxone. This study aims to determine the additional benefit of the extra gram of ceftriaxone when treating SBP in the clinically relevant terms of renal injury and mortality.\n\n\nMethods\n\nThis is a retrospective, single center review of prospectively maintained medical records for all consecutive patients treated with ceftriaxone for SBP at the Beth Israel Deaconess Medical Center, Boston, USA, between January 2003 and December 2011.\n\nWe searched our clinical database for all patients that received ceftriaxone within 48 hours of a peritoneal fluid cell count and differential drawn in the emergency department or hospital ward. We then limited the population to those with 250 or more neutrophils in the ascites. Patient charts were then examined to exclude those with a prior liver transplant, evidence of intra-abdominal source of infection [abscess, perforation, recent (within 2 weeks) intra-abdominal surgery], peritoneal dialysis, or documentation of a secondary infection (urinary tract infection, pneumonia, blood stream infection, cellulitis, meningitis) for which ceftriaxone was started prior to the peritoneal fluid collection. We collected data on age, sex, Model for End Stage Liver Disease (MELD) score (bilirubin, creatinine and PT/INR) at diagnosis, peritoneal white blood cell count and differential, blood and peritoneal culture data, dose of ceftriaxone, additional antibiotics, duration of antibiotic therapy, creatinine trends, intensive care utilization, length of hospital stay and mortality.\n\nStatistics were performed using SAS 9.2 and included student’s t-test, multivariate regression analysis, and log-rank testing/survival analysis where appropriate. P-value of 0.05 was considered significant for all analyses. While no prior studies have examined the effect of ceftriaxone dosing in order to determine study power, prior studies of ceftriaxone for SBP may be instructive. For example, in comparing 2 g ceftriaxone to cefonicid, the in-hospital death rate during therapy was 13% versus 30% which, assuming an alpha of 0.05, a sample size of 91 gives an 80% power13. However, when examining the broader literature on ceftriaxone, regimens of variable duration (5 vs. 10 days) with 30% vs. 35% 30 day mortality would imply that studies require more than 1600 patients for adequate power8–14.\n\n\nResults\n\nWe found 138 patients with SBP treated with ceftriaxone. Of these, 91 patients met our inclusion criteria: 34 patients received 1 g daily and 57 received 2 g (total) daily. There was no significant difference between the groups with respect to age, gender, MELD score, peritoneal culture positivity or other infectious burden (Table 1). All patients had received a protocol of albumin infusion on days 1 and 3 after the diagnosis of SBP in accordance with best practice18.\n\nLooking at patients admitted to a floor service, excluding prior transplants and prior episodes of SBP.\n\nPatients with documented pneumonia or urinary tract infection\n\nMELD = Model for End-Stage Liver Disease. SBP = Spontaneous Bacterial Peritonitis.\n\nWe next compared the hospital course for patients that received either dose of ceftriaxone (Table 2). While both groups were likely to be treated with at least one additional antibiotic during their hospitalization (74% of those treated with 1 g, and 61% of those treated with 2 g), this difference was not significant. The total course of antibiotics – ceftriaxone or otherwise – was also similar between groups. The group receiving 2 g ceftriaxone daily did have a trend towards a shorter hospital stay, although this did not meet statistical significance (13.24 days vs. 10.28, p = 0.44). We did see a statistically significant shorter average length of intensive unit (ICU) stay in patients who received 2 g ceftriaxone a day (0.59 ± 1.78 days), compared to those who received 1 g ceftriaxone daily (3.26 ± 6.9 days) (p = 0.034). The odds ratio for ICU utilization was 0.11 (95% CI 0.02–0.65). However, this difference was no longer significant after controlling for MELD score - odds ratio 0.21 (95% CI 0.04–1.07). Finally, we examined one-year survival for patients treated with 1 versus 2 g ceftriaxone, and found a significant improvement in survival associated with the 2 g dose (p 0.0034 log rank test) (Figure 1). Median overall survival was greater for patients treated with the 2 g dose (228 days vs. 102 days, however it was not significant (p = 0.26).\n\nICU = intensive care unit. Not significant after controlling for MELD score\n\nGiven the high prevalence of additional antibiotic treatment, we also examined the pattern of antibiotic use. Overall, 70% of patients were treated with at least one additional antibiotic. The duration of antibiotic use, as well as the number and type of antibiotics prescribed were highly variable (Table 3). While vancomycin was the most common concurrent antibiotic, used in 46% of patients, over 25 different medications were used in a variety of combinations. To further understand the antibiotic regimens observed, we next examined the available culture data. Of 91 patients diagnosed with SBP on neutrophil criteria, 13 were culture-positive. Of these, one patient had a documented infection resistant to ceftriaxone. This patient was excluded from the analysis. 14 patients had evidence of a secondary infection (Table 1). These included pneumonia (diagnosed with chest x-ray), urinary tract infection (>100,000 colonies on urine dipstick with positive urine culture), and cellulitis (clinical diagnosis documented in chart).\n\n*Recorded as treatment. This analysis excluded patients continued on ciprofloxacin or trimethoprim-sulphamethoxazole for prophylaxis.\n\n\nConclusion\n\nOur study of ceftriaxone dosage for SBP yielded two core findings. First, a total ceftriaxone dose of 2 g daily over 1 g daily exhibited a non-significant reduction of intensive care utilization by cirrhotic patients with SBP, controlling for MELD score. Similarly, there was a trend towards improved mortality with the 2 g dosage. Prospective studies in a larger cohort are indicated to explore the true significance of these results. While it could explain our results, whether the pharmacodynamics of intravenous ceftriaxone are such that the peritoneal drug concentration following a 1 g infusion results in slower control of infection is unclear from our study19. Second, the number, duration and complexity of antibiotic regimens that cirrhotic patients experience is highly variable. The reasons for this finding are unclear and deserve further study in order to understand both the physician and patient factors that increase antibiotic regimen complexity as well as the effect on outcomes including mortality, morbidity and future infection with resistant organisms.\n\nThis study emphasizes the need for antibiotic stewardship and treatment standardization in the care of cirrhotic patients. We feel this can be easily achieved by computer programming. For centers that use electronic provider order entry, a preset dose of 2 g of ceftriaxone when prescribing for a diagnosis of SBP can ensure standardized and appropriate dosing. Our findings are inconclusive but suggestive of a benefit from a higher ceftriaxone dose. As a result, we have programmed a prompt into ceftriaxone orders that asks the physician to specify whether the medication is intended to treat SBP. This selection results in an automatic 2 g daily dose (Figure 2). Cirrhotic patients can be admitted to any service of the hospital, including those staffed by hepatologists, internists, surgeons and intensivists. By standardizing care delivery, the healthcare system can ensure that the medications cirrhotic patients receive are dosed appropriately for their needs. Furthermore, by programming a menu-selection for SBP, our hospital - or any hospital with similar capabilities - may track the disease indications for each antibiotic allowing for audits and outreach.\n\nA. When an ordering physician chooses ceftriaxone, an indication must be chosen. B. When spontaneous bacterial peritonitis is the chosen indication, the preset dose is 2 g daily.\n\nOur conclusions are limited in a few ways. First, our study is retrospective and therefore we cannot comment on the impact of other treatment decisions that may or may not be associated with the dose of ceftriaxone chosen. Additionally, we cannot exclude the possibility that our study was underpowered to detect a difference between treatment groups. Second, the microbiology of our patients’ SBP is unclear given the low rate of culture positivity so we cannot comment on the impact of antimicrobial resistance. Third, follow-up paracenteses to confirm resolution of the SBP after antibiotic treatment were infrequent and thus we cannot comment on the rate of resolution of neutrophilia as function of ceftriaxone dose.\n\nIn order to prevent unwanted practice variation, we recommend standardizing the treatment of SBP by automating the dose of ceftriaxone in the provider order entry system. Further research must be aimed at rationalizing the antibiotic regimens employed in the treatment of cirrhotic patients. Programs to this end include fastidious antibiotic stewardship facilitated by computerized audits of indication-based antibiotic usage and improved microbial culture and detection techniques.\n\n\nData availability\n\nfigshare: Spontaneous bacterial peritonitis outcome and ceftriaxone dosage data, http://dx.doi.org/10.6084/m9.figshare.93175420", "appendix": "Author contributions\n\n\n\nLM: data acquisition, interpretation, drafting of manuscript, statistical analysis. ET: concept of study, data acquisition, writing, analysis. GP: data acquisition, interpretation, revision. ML: critical revision, interpretation. All authors approved this 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\nRunyon BA: AASLD Practice Guidelines Committee: Management of adult patients with ascites due to cirrhosis: an update. Hepatology. 2009; 49(6): 2087–2107. PubMed Abstract \| Publisher Full Text\n\nGinès P, Cárdenas A, Arroyo V, et al.: Management of Cirrhosis and Ascites. N Engl J Med. 2004; 350(16): 1646–1654. PubMed Abstract \| Publisher Full Text\n\nPlanas R, Montoliu S, Ballesté B, et al.: Natural history of patients hospitalized for management of cirrhotic ascites. Clin Gastroenterol Hepatol. 2006; 4(11): 1385–1394. PubMed Abstract \| Publisher Full Text\n\nRimola A, García-Tsao G, Navasa M, et al.: International Ascites Club: Diagnosis, treatment and prophylaxis of spontaneous bacterial peritonitis: a consensus document. J Hepatol. 2000; 32(1): 142–153. PubMed Abstract \| Publisher Full Text\n\nThuluvath PJ, Morss S, Thompson R: Spontaneous bacterial peritonitis--in-hospital mortality, predictors of survival, and health care costs from 1988 to 1998. Am J Gastroenterol. 2001; 96(4): 1232–6. PubMed Abstract \| Publisher Full Text\n\nGinès P, Guevara M, Arroyo V, et al.: Hepatorenal syndrome. Lancet. 2003; 362(9398): 1819–27. PubMed Abstract \| Publisher Full Text\n\nFollo A, Llovet JM, Navasa M, et al.: Renal impairment after spontaneous bacterial peritonitis in cirrhosis: Incidence, clinical course, predictive factors and prognosis. Hepatology. 1994; 20(6): 1495–1501. PubMed Abstract \| Publisher Full Text\n\nCartier M, Terg R, Lucero R, et al.: Pilot study: Gelafundin (polygeline) 4% plus antibiotics in the treatment of high-risk cirrhotic patients with spontaneous bacterial peritonitis. Alimentary Pharm Ther. 2010; 32(1): 43–48. PubMed Abstract \| Publisher Full Text\n\nBaskol M, Gursoy S, Baskol G, et al.: Five days of ceftriaxone to treat culture negative neutrocytic ascites in cirrhotic patients. J Clin Gastroenterol. 2003; 37(5): 403–405. PubMed Abstract\n\nFrança AV, De Souza JB, Silva CM, et al.: Long-term Prognosis of Cirrhosis After Spontaneous Bacterial Peritonitis Treated With Ceftriaxone. J Clin Gastroenterol. 2001; 33(4): 295–298. PubMed Abstract\n\nJavid G, Khan BA, Khan BA, et al.: Short-course ceftriaxone therapy in spontaneous bacterial peritonitis. Postgrad Med J. 1998; 74(876): 592–595. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMesquita MA, Balbino EP, Albuquerque RS, et al.: Ceftriaxone in the treatment of spontaneous bacterial peritonitis: ascitic fluid polymorphonuclear count response and short-term prognosis. Hepatogastroenterology. 1997; 44(17): 1276–80. PubMed Abstract\n\nGómez-Jiménez J, Ribera E, Gasser I, et al.: Randomized Trial Comparing Ceftriaxone with Cefonicid for Treatment of Spontaneous Bacterial Peritonitis in Cirrhotic Patients. Ant Ag Chemo. 1993; 37(8): 1587–1592. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMercader J, Gomez J, Ruiz J, et al.: Use of ceftriaxone in the treatment of bacterial infections in cirrhotic patients. Chemotherapy. 1989; 35(Suppl 2): 23–6. PubMed Abstract \| Publisher Full Text\n\nPham PA, Bartlett JG: Ceftriaxone. Johns Hopkins POC-IT Center. 9/24/2010.\n\nRunyon BA: AASLD Practice Guidelines Committee: Management of adult patients with ascites due to cirrhosis: An update. Hepatology. 2009; 49(6): 2087–2107. PubMed Abstract \| Publisher Full Text\n\nEuropean Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J Hepatol. 2010; 53(3): 397–417. PubMed Abstract \| Publisher Full Text\n\nRunyon BA: AASLD Practice Guidelines Committee: Management of adult patients with ascites due to cirrhosis: an update. Hepatology. 2009; 49(6): 2087–107. PubMed Abstract \| Publisher Full Text\n\nSteib A, Jacoberger B, Von Bandel M, et al.: Concentrations in plasma and tissue penetration of ceftriaxone and ornidazole during liver transplantation. Antimicrob Agents Chemother. 1993; 37(9): 1873–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaura M, Elliot BT, Gail P, et al.: Spontaneous bacterial peritonitis outcome and ceftriaxone dosage data. figshare. 2014. Data Source" }	[ { "id": "3731", "date": "25 Feb 2014", "name": "José Castellote", "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 an interesting retrospective study but I have the following comments:The title is appropriate. I think that the abstract provides an adequate summary but the conclusions may need to be changed. The first sentence is a recommendation and the differences in one-year survival are difficult to explain purely on the basis of 7 days of antibiotic therapy. In the method section the clinical endpoints that the authors are going to study should be explained and the following should be added: in-hospital mortality, infection cure rate, bacterial resistance and super-infections. Ascitic culture method should also be detailed. What are the reasons for different mortality rates at one-year in both groups? This point is crucial and must be discussed in the discussion. Lastly, I think that the study shows no differences in hospital mortality rate, infection control rate, renal failure, or hepatorenal syndrome between both groups and this should be pointed out in the conclusions.", "responses": [ { "c_id": "869", "date": "18 Jun 2014", "name": "Elliot Tapper", "role": "Author Response", "response": "The reviewer’s comments are well taken.The language used in the conclusions has been moderated for less sweeping claims. The use of 1 year mortality after one clinical event is frequently described in the literature on cirrhosis. While SBP and 1 year mortality seem disconnected, SBP is a watershed moment for patients with ascites, the prognostic effects of which are well described. Endpoints have been clarified. Mortality has been included. Infection cure rate is incompletely captured as very few patients received follow up paracentesis; patients with resistant species (very few) and super-infections were specifically excluded from the study. The ascitic culture method is standard. We add a specific comment on the causes of mortality. Please see our response to Dr. Cardenas. We regret that we cannot provide actual causes of death. However, as the primary outcome was all-cause mortality and we controlled for factors that are validated to predict all-cause mortality, we feel that this outcome is legitimate. Agreed." } ] }, { "id": "4389", "date": "08 Apr 2014", "name": "Manuela Merli", "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 a retrospective study aimed at evaluating the relationship between dosing of ceftriaxone (1 or 2 grams) and outcome in SBP. This issue is certainly of interest however, as the authors stated in the discussion, the study is inconclusive due to several limitations which derive from the retrospective approach. For this reason I would also suggest that the title should be changed to underline the point that the main finding in the study is that uneventfully these patients may receive different antibiotic dosages for the same indication. A possible title could be “Need for antibiotic stewardship and treatment standardization in the care of cirrhotic patients”The answer to the question “should we use 1 or 2g ceftriaxone?” can only be derived from a prospective randomized study. On the other hand it is unlikely that one year mortality (Figure 1) could have been influenced by the treatment of the index episode. In fact these patients had a similar length of stay, a similar in hospital mortality and a similar rate of 30 days readmission. From this point of view, one could even derive that the therapy with 1g ceftriaxone was not inferior to 2g ceftriaxone. I suggest these observations be taken into account in the text.", "responses": [ { "c_id": "868", "date": "18 Jun 2014", "name": "Elliot Tapper", "role": "Author Response", "response": "The reviewers’ points are well taken and have been adopted in the manuscript. Language reinforcing the statistically equal efficacy of ceftriaxone doses has been inserted at critical points. The notion that 1g is non-inferior to 2g however cannot be included as the study design does not allow for claims regarding inferiority. Furthermore, we believe that though the power calculation based on prior works suggested our sample size was sufficient, it is likely the case that when comparing 1g to 2g, the study was underpowered to confirm the trends toward improved outcomes with 2g." } ] }, { "id": "3998", "date": "10 Apr 2014", "name": "Andres Cardenas", "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 an interesting retrospective study where the authors review the effectiveness of ceftriaxone dosing for patients with SBP. I like the article and the principles behind the analysis. Major points:The aim of the study should be better explained - what did the authors set out to study? Cure, outcomes, prognosis? The title should reflect the retrospective nature of the analysis. Were any patients on antibiotic prophylaxis (i.e norfloxacion or cipro) prior to the dx of SBP or other bacterial infections? There is a trend of differences in mortality. Why is this so - do the authors have the causes of death in both groups? The authors’ message should be that there seems to be a difference and that this analysis paves the way for future randomized studies that take local microbiological data from each institution into account.", "responses": [ { "c_id": "867", "date": "18 Jun 2014", "name": "Elliot Tapper", "role": "Author Response", "response": "We appreciate Dr. Cardenas' comments. As regards to his major points, the first three were easily addressed. First, the title, abstract and introduction have been changed substantially to clarify the aims with the principle focus being patient outcomes. Second, the title has been changed. Third, we clarified that patients on antibiotic prophylaxis were excluded.As for his fourth, unfortunately it is extremely difficult to determine causes of death when patients often die at other institutions. We do not feel that this is a major limitation for three main reasons. First, we confirm that the patients are alive or dead using a national database. Second, the available prognostics in liver disease, namely the MELD, are capable of predicting all-cause mortality and we adjust for MELD. Third, the causes of death in decompensated cirrhosis are fairly circumscribed and typically closely related to the patients’ antecedent clinical course. SBP can result in mortality via sepsis, and renal failure but also, potentially, variceal bleeding, if say the clinician held beta-blockade to preserve renal function." } ] } ]	1	https://f1000research.com/articles/3-57
https://f1000research.com/articles/3-162/v1	11 Jul 14	{ "type": "Research Article", "title": "Local membrane deformation and micro-injury lead to qualitatively different responses in osteoblasts", "authors": [ "G. Monserratt Lopez-Ayon", "Heng-Yen Liu", "Shu Xing", "Osama M. Maria", "Jeffrey M. LeDue", "Helene Bourque", "Peter Grutter", "Svetlana V. Komarova", "G. Monserratt Lopez-Ayon", "Heng-Yen Liu", "Shu Xing", "Osama M. Maria", "Jeffrey M. LeDue", "Helene Bourque", "Peter Grutter" ], "abstract": "Micro-damage of bone tissue is known to regulate bone turnover. However, it is unknown if individual bone cells can differentiate between membrane deformation and micro-injury. We generated osteoblasts from mouse bone marrow or bone morphogenetic protein 2-transfected C2C12 cells. Single cells were mechanically stimulated by indentation with the atomic force microscopy probe with variable force load either resulting in membrane deformation only, or leading to membrane penetration and micro-injury. Changes in the cytosolic free calcium concentration ([Ca2+]i) in fluo4-AM loaded cells were analyzed. When deformation only was induced, it resulted in an immediate elevation of [Ca2+]i which was localized to the probe periphery. Multiple consecutive local Ca2+ responses were induced by sequential application of low level forces, with characteristic recovery time of ~2 s. The duration of [Ca2+]i elevations was directly proportional to the tip-cell contact time. In contrast, cell micro-injury resulted in transient global elevations of [Ca2+]i, the magnitude of which was independent of the tip-cell contact time. Sequential micro-injury of the same cell did not induce Ca2+ response within 30 s of the first stimulation. Both local and global Ca2+elevations were blocked in Ca2+-free media or in the presence of stretch-activated channel blocker Gd3+. In addition, amount of Ca2+ released during global responses was significantly reduced in the presence of PLC inhibitor Et-18-OCH3. Thus, we found qualitative differences in calcium responses to mechanical forces inducing only membrane deformation or deformation leading to micro-injury.", "keywords": [ "Mechanical stimulation of bone is well-known to regulate bone volume", "structure and composition1", "2. It was recently suggested that", "in addition to deformation forces", "microdamage plays an important role in regulating bone turnover and strength3. Bone is restructured through the coordinated action of bone cells", "osteoblasts responsible for bone formation and osteoclasts responsible for bone resorption. Cells of osteoblastic origin", "including osteoblasts", "osteocytes and bone lining cells are believed to act as mechanosensors in bone tissue1", "4. Understanding how bone cells sense and react to mechanical forces is important for gaining insight into the mechanisms of bone adaptation to its mechanical environment." ], "content": "Introduction\n\nMechanical stimulation of bone is well-known to regulate bone volume, structure and composition1,2. It was recently suggested that, in addition to deformation forces, microdamage plays an important role in regulating bone turnover and strength3. Bone is restructured through the coordinated action of bone cells, osteoblasts responsible for bone formation and osteoclasts responsible for bone resorption. Cells of osteoblastic origin, including osteoblasts, osteocytes and bone lining cells are believed to act as mechanosensors in bone tissue1,4. Understanding how bone cells sense and react to mechanical forces is important for gaining insight into the mechanisms of bone adaptation to its mechanical environment.\n\nBecause of the complexity of the bone environment in vivo, several models have been developed to understand the effects of mechanical stimulation on bone cells in vitro5. These models include application of hydrostatic pressure, longitudinal substrate stretch and fluid shear. These studies have established that osteoblastic cells can sense mechanical stimulation through plasma membrane receptors, stretch activated ion channels, as well as focal adhesion sites6. Ca2+ signaling was shown to be the prominent first response of osteoblastic cells to any type of mechanical stimulation7–9. Ca2+ signaling induced by mechanical stimulation in turn influences numerous bone cell functions such as cytoskeletal reorganization7, gene expression10, proliferation and differentiation6. However, these studies also identified significant complex signaling interactions between multiple cells6, making it difficult to de-convolute the responses of single cells to mechanical stimulation. Moreover, these techniques do not allow exact control of forces applied to individual cells, nor do they report single cell micro-injury.\n\nLocal indentation techniques allow characterization of the response to mechanical stimulation at the single-cell level. Pipette microinjection11, pipette suction11–13, and atomic force microscopy (AFM)14,15 have been used to study responses of individual osteoblasts to mechanical stimulation. From these techniques, only AFM allows application of precisely controlled forces in the nano-Newton scale, as well as provides readout of the extent of membrane deformation16. Moreover, AFM can be used with cantilever tips of different geometries, which allow additional control of the distribution of the force. Spherical tip allows creating high range of membrane strains17, while pyramidal tip allows reversible membrane penetration, which does not result in long-term cell damage18,19.\n\nThe goal of this study was to examine how a single osteoblastic cell responds to forces inducing cell membrane deformation only, or membrane deformation resulting in micro-injury. We used either primary bone marrow cells cultured with ascorbic acid, or C2C12 cells stably transfected with bone morphogenic protein (BMP) 2. C2C12 cells have been shown to undergo osteoblastic differentiation when treated with BMP-220. Mechanical forces of different magnitude were applied using AFM. To monitor cell responsiveness to mechanical forces, changes in cytosolic free Ca2+ concentration ([Ca2+]i) were assessed.\n\n\nMaterials and methods\n\nAll procedures were approved by McGill University’s Animal Care Committee (protocol number 2013-7332) and conformed to the ethical guidelines of the Canadian Council on Animal Care. Six week old male C57/BL6 mice (Charles River) were acclimatized for 1 week, fed ad libitum, and kept on a 12-hour light/dark cycle prior to euthanasia by CO2 asphyxiation followed by cervical dislocation. Femora and tibiae were isolated and separated from soft tissue. The bones were cut in two, placed in an Eppendorf tube, centrifuged twice at 103 rpm for 30 seconds. Cells (~2×107 cells/mouse) were re-suspended in Minimum Essential Medium (MEM (Eagle), from Invitrogen) supplemented with 2 mM of L-glutamine, 100 IU of penicillin, 100 μg/ml of streptomycin and 10% of fetal bovine serum (Wisent), and 5×106 cells were plated on round 25 mm No.1 glass coverslips (Matteck Corporation), and cultured with 50 μg/ml of ascorbic acid at 5% CO2, 37°C for 4–6 days to 50–70% of confluence, which allowed easy identification of individual cells. The media was replaced every third day. The osteoblastic phenotype was confirmed by alkaline phosphatase staining (Fast Red, Sigma).\n\nC2C12 cells stably transfected with BMP-2 (kindly provided by Dr M. Murshed, McGill University) were plated at 2.5×104 cells on round 25 mm No.1 glass coverslips (Matteck corporation). Cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, Invitrogen) supplemented with 2 mM of L-glutamine, 100 IU of penicillin and 100 μg/ml of streptomycin and 10% of fetal bovine serum at 5% CO2, 37°C to 50–70% of confluence. The media was changed every third day. The osteoblastic phenotype was confirmed by alkaline phosphatase staining.\n\nThe cells were loaded with 1.5 μl of Ca2+-sensitive dye fluo4–AM (Molecular Probes, stock solution of 1 mg/ml in DSMO), added to 2 ml of culture media for 40 minutes at room temperature. The cells were washed twice with physiological solution (130 mM NaCl; 5 mM KCl; 1 mM MgCl2; 1 mM CaCl2; 10 mM glucose; 20 mM HEPES, pH 7.4, for Ca2+-containing experiments, or 0 mM CaCl2 and 10 mM EGTA for Ca2+-free experiments)21, the coverslip was assembled onto the peek fluid chamber (Asylum Research) and physiological buffered solution (1.5 ml) was added after loading with fluo4-AM. Gd3+ (Sigma) was dissolved directly in buffer to 50 μM final concentration. 1-O-Octadecyl-2-O-methyl-sn-glycero-3-phosphorylcholine (Et-18-OCH3, Sigma) stock solution (5 mg/ml) was prepared in ethanol, which was used as a vehicle (0.07%) for corresponding experiments. Et-18-OCH3 was used at 5 μM final concentration. Cells were pretreated with inhibitors for 45 min at room temperature before mechanical stimulation. The cells were illuminated with 488 nm laser light and the emitted light was collected with the Cascade II camera. In each experiment a time-sequence of fluorescence images was acquired with one frame taken every 345 ± 10 ms with 250 ms exposure time. Traces were extracted from the video files using a code for Matlab 2011A provided in22. The fluorescence signal in cells not exposed to mechanical stimulation did not demonstrate associated changes and was used as a reference for signal correction for bleaching and cantilever reflection artifacts using the protocol for data processing for systematic signal recovery that we established and described previously22.\n\nThe experiments were conducted using an MFP-3D-BIO AFM (Asylum Research, Santa Barbara CA) mounted on an Olympus IX-71 inverted optical microscope. The sample placed in the closed fluid cell was left undisturbed for 15 min to achieve thermal equilibrium at 37°C. A 60X oil immersion objective with 1.45 NA (Olympus) was put into contact with the coverslip allowing optical access from the bottom and AFM access on top of the sample. The region of interest was located and aligned with the cantilever tip using the bright field and fluorescence images.\n\nThe cantilever tip (NCLAuD, Nanosensors), was etched down to 1 μm2 contact area with a focused ion beam microscope (FEI DB235). When sharp tips were used, the membrane rupture force event was not distinguishable in the force curve. Using Hooke’s law (F=k⋅d), the force exerted by the probe (F) was determined from the cantilever spring constant (k, 39 ± 7.8 N/m) and deflection (d). The net extension of the piezo element, the speed, the time of indentation, the time between indentations and the number of indentations were controlled using the Asylum MFP3D software in IgorPro 6.22 platform.\n\nData are presented as representative images, representative traces, means ± SE with n being the number of experiments analyzed. The normalized amplitude as a function of time, the amplitude decay rate as a function of recovery time and the response duration as a function of the contact time were analyzed using least squares regression and fitted a line or an exponential as appropriate. Categorical data were analyzed as described previously23. Statistical differences were assessed using Fisher Exact Probability test for categorical data or Student’s t-test for continuous data and were accepted as significant at p < 0.05. Statistical analysis was performed in Microsoft Excel 2007.\n\n\nResults\n\nWe employed two cell models to study osteoblast mechanosensitivity: i) mouse bone marrow cells cultured in the presence of ascorbic acid (50 mg/ml) for 4–6 days (Figure 1A) or ii) C2C12 cells stably transfected with BMP-2 and cultured for 2–6 days (Figure 1B). Both models represent osteoblastic cells at the early differentiation stage due to the limitation of AFM use in confluent and multilayered cultures. Osteoblastic phenotype was confirmed in fixed cultured by alkaline phosphatase staining. In parallel live cultures, the cells exhibiting osteoblastic morphology (strongly adhered cells with relatively large body and several filopodia) were chosen for mechanical testing. Mechanical stimulation was performed with AFM force spectroscopy indentation (Figure 1C) using a 1 μm2 area tip (Figure 1D). Using AFM allows strict control of the amount of force applied to the cell (Figure 1E, F). The membrane penetration event is easily identifiable on the force-distance curve as a decrease in the force required to continue to move the probe. The membrane penetration force was found to be similar for osteoblasts obtained in different cultures: 516 ± 200 nN for primary osteoblasts and 672 ± 100 nN for C2C12 osteoblasts. We analyzed two distinct modes of mechanical stimulation: i) the maximum force was set below the membrane penetration force (at ∼400 nN), leading only to membrane deformation (low-load, Figure 1E, F, blue); and ii) the maximum force was set above the membrane penetration force (∼2800 nN), inducing membrane rupture and penetration (high-load, Figure 1E, F, red).\n\nBone marrow cells were cultured in the presence of ascorbic acid (50 μg/ml) for 4–6 days to induce osteoblast differentiation or C2C12 cells stably transfected with BMP-2 were cultured for 2–6 days to obtain osteoblastic phenotype. A, B) Representative images of bone marrow culture on day 5 (A) and C2C12 culture on day 3 (B) stained for osteoblastic marker, alkaline phosphatase (red). C) Schematic representation of the AFM operation under liquid: the sample and the tip are submerged and changes in the laser reflection signal due to deflection of the cantilever are monitored. D) Experiments were performed using a probe with 1 μm2 tip area; the tip is etched using focused ion beam. E) Schematic representation of events observed using force-distance curve: a) as the AFM tip approaches the cell, the contact point is evident as increase in force required to move the probe; b) when the probe indents the cell membrane, the force increases until it reaches the membrane rupture force and the tip penetrates the cell (apparent by a decrease in the cell resistance); c) indentation is continued until a predetermined maximum force is reached, and the probe retraction is initiated. The retraction curve commonly deviates from the approach curve, reflecting that energy is required for deformation and penetration of the cell membrane. The maximum force was set to be either below (blue) or above (red) the membrane rupture force. F) The example force-distance curves from the experiments in which deformation only (blue) or deformation plus penetration (red) were induced.\n\nSince [Ca2+]i elevations are known to be the most common first responses to mechanical stimulation4,9,11,24, we analyzed the changes in [Ca2+]i in response to different loading regimes in the osteoblasts loaded with [Ca2+]i fluorescent indicator fluo4-AM. In the low load indentation regime, where the force exerted on the cell induced membrane deformation (Figure 2A), only local [Ca2+]i transients were observed (Figure 2B). To analyze the transients, the spatially-averaged intensity over a circle of 5 μm in diameter centered on the point of indentation (red circle in Figure 2B) was normalized to the initial baseline signal and plotted as a function of time (Figure 2C). When larger circles were taken, the change in averaged intensity was smaller; if the whole cell was selected, the local response was not apparent. When consecutive low load indentations were performed at the same location on the same cell multiple local responses were induced (Figure 2A, C).\n\nOsteoblasts generated from C2C12 cells were loaded with Ca2+-sensitive dye fluo4-AM and changes in [Ca2+]i in response to mechanical stimulation were assessed. A–C) The maximum force was set below the membrane rupture force. A) Representative force-distance curves depicting multiple local indentations of the cell membrane. B) A single cell exhibiting local [Ca2+]i elevation in response to membrane deformation. Magnified below is the region of the cell centered at the indentation point, that demonstrates the changes in [Ca2+]i with time following a single AFM indentation at 5 s. In pseudo-color, black/blue represents low and yellow/white represents high [Ca2+]i levels. C) Representative trace depicting changes in fluorescence intensity in the region shown in red on B, in response to multiple membrane deformations. D–G) The maximum force was set above the membrane rupture force. D) Force-distance curve demonstrating the membrane deformation and penetration. E) Micrographs demonstrating global increase in fluorescence intensity following a single high-load indentation at 6 s (same color scale as in B). F) Changes in fluorescence intensity in the whole cell shown in E following single indentation with the penetration of cell membrane (indicated by an arrow). G) Magnified region of the cell centered at the high-load indentation point, demonstrating the transient character of membrane damage.\n\nIn the high load regime, which resulted in local membrane rupture and penetration of the cell (Figure 2D), whole cell [Ca2+]i elevations were observed (Figure 2E). When the normalized fluorescence intensity was averaged over an ellipse surrounding the whole cell and plotted as a function of time (Figure 2F), it was apparent that global [Ca2+]i elevations were transient. When we focused on the membrane level to directly visualize the closure of the membrane micro-injury, we observed that it seals within 40–60 s after the stimulation (Figure 2G). It is conceivable that high load indentation may affect the cell viability. However, in all our experiments, the cell previously exposed to high load indentation maintained basal calcium levels within normal range for the whole period of observation (5–30 min), and when directly tested using Trypan Blue exclusion test, six out of six cells previously indented were viable ∼10 minutes after high load indentation. Taken together, these data strongly suggest that the micro-injury induced by AFM probe is reversible and does not result in cell death.\n\nSince application of AFM may potentially interfere with fluorescence measurements, we have previously performed in depth analysis of the system performance22. We have found that in experiments where local repetitive stimulations were performed, reflection artifacts correlated with cantilever motion represented a significant component of the fluorescent signal. We developed a protocol to correct the fluorescence traces for reflection artifacts, as well as photobleaching22. The traces were processed as follows: 1) the region of interest was selected in the digital image; 2) the average fluorescence intensity data were extracted and normalized to the initial basal reading to correct for differences in dye loading; 3) the signal was corrected for bleaching and reflection artifacts as described in22; 4) the noise was determined as the standard deviation in a linear portion of the trace, and 5) the [Ca2+]i response was deemed positive when an increase in intensity exceeded four-fold the average noise. Cells exhibiting spontaneous [Ca2+]i fluctuations25 were excluded from the study.\n\nFirst, we compared the responses in C2C12-derived and primary osteoblasts. We have found that the low load indentation induced qualitatively similar responses in the primary and C2C12-derived osteoblasts (Figure 3A). However, when we compared the percentage of cells exhibiting local [Ca2+]i transients in response to indentation performed at low and high speeds, we found that a higher percentage of C2C12-derived osteoblasts responded to the mechanical stimulation compared to primary osteoblasts (Figure 3B). In addition, a higher percentage of C2C12-derived osteoblasts exhibited [Ca2+]i transients when indented at lower speed compared to higher speed (Figure 3B). In primary osteoblasts a similar trend was observed, however it did not reach statistical significance since the rate of response was low in these cells. We analyzed in depth the [Ca2+]i responses induced in C2C12-derived osteoblasts by a single low-load indentation performed at different speed (2, 10 and 20 μm/s). The [Ca2+]i elevations were further characterized by measuring i) the relative amplitude of the response (with respect to basal); ii) the duration of the response (the width of a transient at the half maximum amplitude); and iii) the amount of [Ca2+]i (duration multiplied by amplitude). The speed of indentation did not significantly affect the amplitude of calcium responses (Figure 3C). However, indentations performed at low speed induced calcium responses of longer duration (Figure 3D) and thus resulted in release of greater amount of [Ca2+]i (Figure 3E) compared to indentations performed at high speed. To analyze if indenting at different speeds delivered different mechanical stimulation, the force curves were characterized by quantifying i) the extent of membrane deformation (the distance between the contact point and the maximum piezo extension) (Figure 3F), and ii) the energy spent to deform the cell (the area between the approach and retraction curves) (Figure 3F, right scale). No statistically significant difference was observed in these parameters when different indentation speeds were compared.\n\nPrimary or C2C12 derived osteoblasts were loaded with fluo4-AM and single indentations were performed at different speeds (2, 10, 20 mm/s). The maximum force was set below the membrane rupture force. A) The examples of local Ca2+ elevations induced by a low level deformation in a primary osteoblast (left) and a C2C12-derived osteoblast (right). B) The percentage of primary (white) or C2C12-derived (black) osteoblasts responding to deformations with local Ca2+ elevations. Data are means ± SE, n = 15 for primary osteoblasts, n=10 and 30 for C2C12 osteoblasts deformed at 2 and 20 μm/s respectively, # indicates p<0.05 difference between primary and C2C12-derived osteoblasts, * indicates p<0.05 difference between deformation speeds, assessed by Fisher Exact Probability test for categorical data. C–E) Local Ca2+ transients in C2C12-derived osteoblasts were analyzed for the amplitude of Ca2+ response (C), the duration of Ca2+ response (D) and the amount of Ca2+ released during the response (E). Data are means ± SE; n is the number of responses from 10 trials: for 2 µm/s n = 6, for 10 µm/s n = 2, for 20 µm/s n = 6, indicates p<0.05 difference as assessed by Student’s t-test. F) Force-distance curves were analyzed and the extent of membrane deformation (black, scale on the left) and energy spent to deform the membrane (white, scale on the right) were assessed. Data are means ± SE, n = 5, no significant difference.\n\nWe next assessed how the contact time (the net time the probe spent in contact with the cell during stimulation) affects calcium responses to low-load indentations performed at different speeds. We varied the contact time by increasing by 1 s the time that the tip spent in the maximally extended piezo position. The reflection of a cantilever on the calcium fluorescence recording allows achieving exact temporal overlay between the fluorescence intensity and the force-distance curve (as described in22). The [Ca2+]i signal was initiated when the tip came into contact with the cell and was maximum at the maximum deflection (Figure 4A, B). When the contact time was increased by 1 s, the comparable increase in calcium response duration was observed (Figure 4C). The mean of the response duration was plotted as a function of the corresponding contact time demonstrating that the average calcium response duration was directly proportional to the contact time, R2 = 0.90 (Figure 4D). We next assessed if calcium responses of different durations can be induced in a single cell by varying the contact time (as the piezo extension time) in consecutive indentations (Figure 4E). We have found that a single cell exhibited [Ca2+]i elevations of different durations (Figure 4E, i-iv, black) when indentations with different contact times (Figure 4E, i-iv, red) were performed.\n\nC2C12-derived osteoblasts were loaded with fluo4-AM and responses to single indentation were assessed. The maximum force was set below the membrane rupture force. A–C) Examples of Ca2+ transients in black (left scale) in response to indentation performed with different contact time. Overlaid are the cantilever deflection curves in red (right scale). The indentation speed was set to 20 mm/s (A) or 2 mm/s (B, C). For C, the probe was maintained in contact with the cell for additional 1 s. D) The Ca2+ transient duration is plotted as a function of contact time. Data are means ± SE, for 0.1 s and 1.2 s n = 4, for 0.2 s n = 2, for 1.0, 1.1 and 2.0 s n = 3. Linear fit is f(x) = 0.98x + 0.60 with R2=0.90. E) A single cell exhibited [Ca2+]i elevations of different duration in response to indentations with different contact times. The indentation speed was 20 µm/s, the probe was maintained in contact with the cell for additional i) 0 s, ii) 0.3 s, iii) 0.7 s and iv) 3 s.\n\nSince we have found that repeated low level mechanical stimulations can elicit multiple consecutive responses in a single cell, we next examined how the exposure to mechanical stimulation affects subsequent responses of the same cell. Multiple low load indentations were performed on the same cell with different frequencies, and [Ca2+]i changes in response were assessed (Figure 5). We have found that when cells were stimulated with higher frequency, the amplitude of the response in consecutive stimulations decreased (Figure 5A–C). Within each sequence of Ca2+ responses to the consecutive indentations, we normalized the amplitudes with respect to the amplitude of Ca2+ response to the first indentation and quantified the amplitude decay rate as a slope (αi) of the best fit line for the responses. This slope is negative, representing a decrease in the amplitude, and when multiplied by 100, this slope is expressed as percentage decay. We have found that amplitude decay is significantly higher when cells are stimulated with higher frequencies (Figure 5D). At different frequencies two parameters are changing – the contact time, defined as the time the probe spends in contact with the cell, and recovery time, defined as the time off contact between stimulations. We next specifically varied the recovery time by introducing a delay of 1 s between consecutive stimulations, and plotted the average amplitude decay rates as a function of the recovery time (Figure 5E). An exponential curve with a characteristic recovery time of 2.00 ± 0.08 s was found to be the best fit for the relationship between the amplitude decay rate and the recovery time. These data suggest that a refractory period of ~10 s is required for complete amplitude recovery after low load mechanical stimulation.\n\nC2C12-derived osteoblasts were loaded with fluo4-AM and changes in [Ca2+]i in response to multiple indentations were assessed. The maximum force was set below the membrane rupture force. A–C) Examples of multiple Ca2+ transients in response to consecutive indentations performed at 10 µm/s, 0.5 Hz (A), 2 µm/s, 0.1 Hz (B) and 2 µm/s, with a 2 s delay between stimulations, 0.08 Hz (C). The amplitudes of the consecutive responses were normalized with respect to the first response and the amplitude decay rate αi was determined as the slope of a linear fit to the amplitude-time graph (dashed line). D) The amplitude decay rates were plotted as a function of the frequency of indentation. Data are means ± SE (except for 0.5 Hz), n is a number of responses from 5 trials: for 0.1 Hz n = 4, for 0.5 Hz n = 1, for 1 Hz n = 4, indicates p<0.05 difference compared to 1 Hz, assessed by Student’s t-test. E) The amplitude decay rates were plotted as a function of the recovery time. Data are means ± SE (except for 4.8 and 5.8 s), n is a number of responses from 5 trials: for 0.9 s n = 3, for 1.9 s n = 4, for 4.8 s and 5.8 s n = 1, for 9 s n = 4 and for 10 s n = 3. The solid line represents an exponential fit: f(x) = a exp(-x b)+c with a time constant of b = 0.5s-1; τ~2s, RMSE = 0.02 s-1 (R2=0.77).\n\nWhen the maximum force was set at 2800 nN (three-fold above the membrane penetration force), the force-distance curves confirmed that membrane penetration occurred in all the cells tested. We have found that qualitatively similar responses were induced in primary and C2C12-derived osteoblasts (Figure 6A, B), however the response rate was significantly lower for the primary osteoblasts (Figure 6C). When the whole cell average fluorescence intensity and the force-distance curves of individual cells were overlaid in time, we found that, in contrast to low-load indentation, micro-injury-induced global [Ca2+]i elevations were delayed with respect to the maximum deflection of the probe (which indicates maximum cell penetration), and reached their amplitude after the tip was released from cell contact (Figure 6A, B). We assessed [Ca2+]i changes induced in C2C12-derived osteoblasts by a single high-load indentation performed at different speeds (2, 10 and 20 μm/s). There was no significant correlation between the indentation speed and the amplitude (Figure 6D), duration (Figure 6E) or amount of Ca2+ released (Figure 6F) during the responses. The force-distance curve analysis did not demonstrate statistically significant differences in the extent of membrane deformation prior to the membrane penetration (Figure 6G), extent of membrane penetration (Figure 6H), energy spent in deformation, micro-injury and penetration (Figure 6I) or membrane rupture force (Figure 6J).\n\nPrimary mouse osteoblasts or osteoblasts generated from C2C12 cells were loaded with fluo4-AM and single indentations were performed. The maximum force was set above the membrane rupture force. A, B) Examples of Ca2+ transients (left scale) in primary osteoblast (A) or C2C12-derived osteoclast (B) in response to the high-load indentations. Overlaid are the cantilever deflection curves in red (right scale). C) The percentage of cells that exhibit global Ca2+ elevations in primary (white) or C2C12-generated (black) osteoblasts in response to the high-load indentations performed with different speed of 2 or 20 μm/s. Data are means ± SE, n = 25 for primary osteoblasts and n=10 for C2C12-derived osteoblasts, # indicates a p<0.05 difference between primary and C2C12-generated osteoblasts as assessed by Fisher Exact Probability test. D–F) Changes in global Ca2+ in response to membrane penetration in C2C12-derived osteoblasts were analyzed for the amplitude of Ca2+ response (D), the duration of Ca2+ response (E) and the amount of Ca2+ released during the response (F). Data are means ± SE, n is the number of response from 10 trials: for 2 µm/s n = 9, for 10 µm/s n = 10, and for 20 µm/s n = 9, no significant difference. G–J) Force-distance curves were analyzed for the extent of membrane deformation (G), penetration depth (H), energy spent to deform and penetrate the membrane (I) and membrane rupture force (J). Data are means ± SE, n = 10, no significant difference.\n\nSince the rise in [Ca2+]i can be due to influx from the extracellular space as well as to release from intracellular stores, we next investigated the contribution of these pathways to membrane deformation and to micro-injury-induced [Ca2+]i responses. We analyzed changes in [Ca2+]i in response to low-load and high-load mechanical stimulations in C2C12 osteoblasts maintained in the control Ca2+-containing buffer or in Ca2+-free, EGTA (10 mM)-containing buffer (Figure 7A–C). Both local [Ca2+]i elevations (Figure 7A, B) and global [Ca2+]i responses (Figure 7C) were prevented by the lack of calcium in the extracellular space. We next inhibited mechanosensitive membrane channels using Gd3+ (50 μM)17. In the presence of Gd3+ both local [Ca2+]i transients in response to membrane deformation (Figure 7D) and micro-injury-induced global [Ca2+]i responses (Figure 7E) were prevented. To inhibit the phospholipase C (PLC)-inositol triphosphate (IP3) pathway leading to calcium release from intracellular stores, we used PLC inhibitor Et-18-OCH3 (5 μM)17. In the low-load regime, local [Ca2+]i transients (Figure 7F) occurred at the same frequency as in control (Figure 7G), and the amount of Ca2+ released during the responses was not significantly affected (Figure 7H). However, global [Ca2+]i transients induced by the high-load stimulation were noticeably smaller in Et-18-OCH3-treated, but not in vehicle-treated cells compared to control (Figure 7I). The percentage of cells exhibiting global [Ca2+]i transients in response to membrane penetration was not significantly affected by Et-18-OCH3 or vehicle (Figure 7J). However, the amount of Ca2+ released during the response was significantly decreased in cells treated with Et-18-OCH3 (Figure 7K). These data indicate that mechanosensitive Ca2+ channels mediate both local and global responses, while the intracellular Ca2+ stores are important for global responses only.\n\nC2C12-generated osteoblasts were loaded with fluo4-AM, and single indentations were performed at 2 μm/s. A) Average local Ca2+ transients in response to low-load indentation in Ca2+-containing (blue) and Ca2+-free (10 mM EGTA, red) buffer. Data are means ± SE. B, C) Percentage of cells exhibiting local (B) and global (C) Ca2+ elevations in Ca2+-containing (+Ca2+) or Ca2+ free (-Ca2+) buffer in response to low-load (B) or high-load (C) indentation. Data are means ± SE, for (B) n = 30 for Ca2+-containing, n = 10 for Ca2+-free; for (C) n = 10; indicates a p<0.05 difference as assessed by Fisher Exact Probability test. D) Average local Ca2+ transients in response to membrane indentation in control (blue) or mechanosensitive Ca2+ channel blocker Gd3+ (50 µM)-containing buffer (red). Data are means ± SE. E) Average global Ca2+ transients in response to membrane penetration in control (blue) or Gd3+-containing buffer (red). Data are means ± SE. F) Average local Ca2+ responses to membrane indentation in control (blue), vehicle (0.07% ethanol)-containing (green) or PLC inhibitor Et-18-OCH3-containing (red) buffer. Data are means ± SE. G) The percentage of cells that exhibited local Ca2+ elevations in response to membrane indentation in control, vehicle- or Et-18-OCH3-containing buffer. Data are means ± SE, n = 30 for control, n = 10 for vehicle and Et-18-OCH3-treated cells, no significant difference. H) Average amount of Ca2+ released in response to membrane indentation in control, vehicle- or Et-18-OCH3-containing buffer. Data are means ± SE, n = 4–5, no significant difference. I) Average global Ca2+ elevations in response to membrane penetration in control (blue), vehicle-containing (green) or Et-18-OCH3-containing (red) buffer. Data are means ± SE. J) The percentage of cells that exhibited global Ca2+ elevations in response to membrane penetration. Data are means ± SE, n = 33 for control, n = 10 for vehicle and Et-18-OCH3-treated cells, no significant difference. K) Average amount of Ca2+ released in response to membrane penetration. Data are means ± SE, n = 8 for control and vehicle, n = 5 for Et-18-OCH3-treated cells, indicates a p<0.05 difference as assessed by Student’s t-test.\n\n\nDiscussion\n\nIn this study, we have examined the ability of individual osteoblasts to respond to mechanical stimulation applied using AFM. Variable force load was applied resulting in either membrane deformation only or in membrane penetration and micro-injury, and changes in [Ca2+]i in fluo-4 loaded cells were analyzed. Qualitatively different Ca2+ responses were observed in different force loads. In response to membrane deformation only, immediate local [Ca2+]i elevations limited to the indentation region were observed. Multiple stimulations of a single cell resulted in Ca2+ responses of similar amplitude if a recovery time of more than 2 s between the stimulations was allowed. The duration of Ca2+ responses to low-load indentation was proportional to the duration of deformation. In contrast, micro-injury induced global Ca2+ elevation, which continued to develop after the removal of the probe and was independent of the duration of indentation. These data demonstrate that Ca2+ responses to local membrane deformation exhibit threshold properties when micro-injury is induced.\n\nIn our model, micro-injury was required in order to achieve global elevations of [Ca2+]i in the majority of cells. Many techniques used to study cell mechanosensitivity, such as fluid flow, substrate strain and pipette indentation, do not allow sufficient resolution to detect membrane micro-injury. However, a number of studies, in which the absence of membrane penetration can be reliably confirmed, demonstrated that global Ca2+ responses can be induced by membrane deformation only11,17. Several differences in the experimental setup could account for this discrepancy. Charras and Horton17 used AFM with spherical tip of 10–30 μm diameter, which would deform a 100-fold larger area compared to the pyramidal tip with 1 μm2 contact area used in our study. Thus, the extent of horizontal membrane involvement in the deformation may be important for cell mechanosensitivity. While Xia and Ferrier11 used patch-clamp micropipette with similar dimensions to the pyramidal tip used in our study, it has been shown that pipette suction creates substantially larger strains compared to AFM micro-indentation14. Therefore, higher calcium responses observed in the Xia and Ferrier study compared to our study are in keeping with previously demonstrated critical role of vertical membrane deformation in cell mechanosensitivity17. We have now shown that the duration of Ca2+ responses is directly proportional to time the cell membrane spends in deformation. Therefore, it is likely that cell sensitivity to membrane deformation is related to a combined effect of: i) the extent of the horizontal involvement of the membrane, ii) the vertical deformation of the membrane and iii) the duration of the deformation. Mechanosensitive calcium channels were strongly implicated in generating Ca2+ responses to mechanical stimulations in previous studies11,17,25 and were confirmed to provide a critical contribution to both local and global Ca2+ responses in our study. Therefore, it is possible that local intracellular Ca2+ acts as an integrating signal that increases when open channels are more numerous or are activated longer, until it reaches the threshold Ca2+ concentration necessary to induce global Ca2+ response. Micro-injury then contributes to increasing local Ca2+ levels to threshold concentrations, inducing global response. Once global calcium elevation is triggered, it proceeds independently of the level of mechanical stimulation, as suggested by the lack of correlation between the magnitude of global Ca2+ response and mechanical stimulation in our study and that of Charras and Horton17. Our data demonstrate that calcium release from intracellular stores plays an important role in this process.\n\nWe identified the refractory periods during which the responses to subsequent mechanical stimulations were either absent or diminished in amplitude. This refractory period following membrane deformation only was relatively short, and the calcium response was fully reestablished after the recovery time of 10 s. In contrast, in experiments where micro-injury and global elevation of calcium were observed, the recovery period was longer than 30 s. Previously, the refractory periods for global Ca2+ responses to fluid flow were reported to be in the order of 600–900 s26. The presence of refractory periods is of potential importance, since it was shown previously, that in order to induce potent osteogenic response, mechanical loading of bone should be performed as a series of repeated loading periods separated by the periods of rest27,28. Of interest, the rest period of 10 s was found to be sufficient to induce potent bone formation in response to low magnitude mechanical loading28,29.\n\nTaken together, our study provides new insights into the complex dynamics of cellular responses to mechanical stimulations. In contrast to many previous studies of cell responses to mechanical forces, atomic force microscopy allows very precise control and monitoring of the physical parameters of the experiment, such as forces and deformations applied at a single cell level. Using a well-established readout of cellular response, calcium signaling, allowed us to identify novel correlates between mechanical stimulation and cell responses. Such knowledge is important for better understanding of the mechanisms of mechanical loading-induced bone formation, as well as micro-damage induced bone remodeling.\n\n\nData availability\n\nfigshare: Data on the effects of local membrane deformation and micro-injury in osteoblasts, doi: http://dx.doi.org/10.6084/m9.figshare.109135630", "appendix": "Author contributions\n\n\n\nGMLA Study conception and design; Acquisition of data; Analysis and interpretation of data; Drafting of manuscript; Critical revision\n\nH-YL Acquisition of data; Analysis of data; Drafting of manuscript\n\nSX Acquisition of data; Analysis of data\n\nOMM Study conception; Sample preparation, preliminary data acquisition\n\nJML Acquisition of data; Analysis and interpretation of data\n\nHB Analysis and interpretation of data\n\nPG Analysis and interpretation of data; Critical revision\n\nSVK Study conception and design; Analysis and interpretation of data; Drafting of manuscript; Critical revision\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by Natural Sciences and Engineering Research Council of Canada Discovery grants to PHG (RGPIN-223110) and SVK (RGPIN-288253).\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 Dr. Gulzhakhan Sadvakassova for preparing and characterizing the cells used in this study, Dr Monzur Murshed, McGill University for the C2C12-BMP-2 cell line. The authors thank DJ Oliver for helpful discussions. GMLA was supported by the Principal’s Graduate Fellowship, Chalk-Rowles Fellowship, NSERC Graduate Excellence Fellowship and McGill University. HYL was funded by NSERC Undergraduate Science Research Award. SVK holds Canada Research Chair.\n\n\nReferences\n\nRubin CT: Skeletal strain and the functional significance of bone architecture. Calcif Tissue Int. 1984; 36(Suppl 1): S11–8. PubMed Abstract \| Publisher Full Text\n\nSuva LJ, Gaddy D, Perrien DS, et al.: Regulation of bone mass by mechanical loading: microarchitecture and genetics. Curr Osteoporos Rep. 2005; 3(2): 46–51. PubMed Abstract \| Publisher Full Text\n\nBurr D: Microdamage and bone strength. Osteoporos Int. 2003; 14(Suppl 5): S67–72. PubMed Abstract \| Publisher Full Text\n\nBurger EH, Klein-Nulend J: Mechanotransduction in bone--role of the lacuno-canalicular network. FASEB J. 1999; 13(Suppl): S101–12. PubMed Abstract\n\nBrown TD: Techniques for mechanical stimulation of cells in vitro: a review. J Biomech. 2000; 33(1): 3–14. PubMed Abstract \| Publisher Full Text\n\nLiedert A, Kaspar D, Blakytny R, et al.: Signal transduction pathways involved in mechanotransduction in bone cells. Biochem Biophys Res Commun. 2006; 349(1): 1–5. PubMed Abstract \| Publisher Full Text\n\nChen NX, Ryder KD, Pavalko FM, et al.: Ca(2+) regulates fluid shear-induced cytoskeletal reorganization and gene expression in osteoblasts. Am J Physiol Cell Physiol. 2000; 278(5): C989–97. PubMed Abstract\n\nHung CT, Allen FD, Pollack SR, et al.: Intracellular Ca2+ stores and extracellular Ca2+ are required in the real-time Ca2+ response of bone cells experiencing fluid flow. J Biomech. 1996; 29(11): 1411–7. PubMed Abstract \| Publisher Full Text\n\nHung CT, Pollack SR, Reilly TM, et al.: Real-time calcium response of cultured bone cells to fluid flow. Clin Orthop Relat Res. 1995; (313): 256–69. PubMed Abstract\n\nPapachristou DJ, Papachroni KK, Basdra EK, et al.: Signaling networks and transcription factors regulating mechanotransduction in bone. Bioessays. 2009; 31(7): 794–804. PubMed Abstract \| Publisher Full Text\n\nXia SL, Ferrier J: Propagation of a calcium pulse between osteoblastic cells. Biochem Biophys Res Commun. 1992; 186(3): 1212–9. PubMed Abstract \| Publisher Full Text\n\nKirber MT, Guerrero-Hernandez A, Bowman DS, et al.: Multiple pathways responsible for the stretch-induced increase in Ca2+ concentration in toad stomach smooth muscle cells. J Physiol. 2000; 524(Pt 1): 3–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZou H, Lifshitz LM, Tuft RA, et al.: Visualization of Ca2+ entry through single stretch-activated cation channels. Proc Natl Acad Sci U S A. 2002; 99(9): 6404–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCharras GT, Williams BA, Sims SM, et al.: Estimating the sensitivity of mechanosensitive ion channels to membrane strain and tension. Biophys J. 2004; 87(4): 2870–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuo XE, Takai E, Jiang X, et al.: Intracellular calcium waves in bone cell networks under single cell nanoindentation. Mol Cell Biomech. 2006; 3(3): 95–107. PubMed Abstract \| Publisher Full Text\n\nLamontagne CA, Cuerrier C, Grandbois M: AFM as a tool to probe and manipulate cellular processes. Pflugers Arch. 2008; 456(1): 61–70. PubMed Abstract \| Publisher Full Text\n\nCharras GT, Horton MA: Single cell mechanotransduction and its modulation analyzed by atomic force microscope indentation. Biophys J. 2002; 82(6): 2970–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCuerrier CM, Lebel R, Grandbois M: Single cell transfection using plasmid decorated AFM probes. Biochem Biophys Res Commun. 2007; 355(3): 632–636. PubMed Abstract \| Publisher Full Text\n\nVakarelski IU, Brown SC, Higashitani K: Penetration of living cell membranes with fortified carbon nanotube tips. Langmuir. 2007; 23(22): 10893–10896. PubMed Abstract \| Publisher Full Text\n\nKatagiri T, Yamaguchi A, Komaki M, et al.: Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J Cell Biol. 1994; 127(6 Pt 1): 1755–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKemeny-Suss N, Kasneci A, Rivas D, et al.: Alendronate affects calcium dynamics in cardiomyocytes in vitro. Vascul Pharmacol. 2009; 51(5–6): 350–8. PubMed Abstract \| Publisher Full Text\n\nLopez-Ayon GM, Oliver DJ, Grutter PH, et al.: Deconvolution of calcium fluorescent indicator signal from AFM cantilever reflection. Microsc Microanal. 2012; 18(4): 808–15. PubMed Abstract \| Publisher Full Text\n\nXu B, Feng X, Burdine RD: Categorical data analysis in experimental biology. Dev Biol. 2010; 348(1): 3–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDuncan RL, Turner CH: Mechanotransduction and the functional response of bone to mechanical strain. Calcif Tissue Int. 1995; 57(5): 344–58. PubMed Abstract \| Publisher Full Text\n\nNishitani WS, Saif TA, Wang Y: Calcium signaling in live cells on elastic gels under mechanical vibration at subcellular levels. PLoS One. 2011; 6(10): e26181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDonahue SW, Donahue HJ, Jacobs CR: Osteoblastic cells have refractory periods for fluid-flow-induced intracellular calcium oscillations for short bouts of flow and display multiple low-magnitude oscillations during long-term flow. J Biomech. 2003; 36(1): 35–43. PubMed Abstract \| Publisher Full Text\n\nRobling AG, Hinant FM, Burr DB, et al.: Improved bone structure and strength after long-term mechanical loading Is greatest if loading Is separated into short bouts. J Bone Miner Res. 2002; 17(8): 1545–1554. PubMed Abstract \| Publisher Full Text\n\nSrinivasan S, Weimer DA, Agans SC, et al.: Low-magnitude mechanical loading becomes osteogenic when rest is inserted between each load cycle. J Bone Miner Res. 2002; 17(9): 1613–1620. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSrinivasan S, Agans SC, King KA, et al.: Enabling bone formation in the aged skeleton via rest-inserted mechanical loading. Bone. 2003; 33(6): 946–955. PubMed Abstract \| Publisher Full Text\n\nLopez-Ayon GM, Liu HY, Xing S, et al.: Data on the effects of local membrane deformation and micro-injury in osteoblasts. figshare. 2014. Data Source" }	[ { "id": "5433", "date": "06 Aug 2014", "name": "Núria Gavara", "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've read with interest the manuscript by Lopez-Ayon at el. Overall, I think the authors have done a good job in characterizing qualitatively the differences in Ca2+ responses induced by local membrane deformation and micro-injury. I particularly liked the measurements involving cyclic membrane perturbation and inhibition of Ca2+ channels and internal store. I think the latter provide a good understanding on the different origin of the Ca2+ released when cells are probed or micro-injured. Nevertheless, i have a technical concern and some questions that i feel the authors could have discussed more in-depth in the manuscript:Figure 4c and 4e(iv) show large deviations in the deflection of the cantilever when cells are indented for longer time periods. Which kind of feedback was used to control the AFM tip position? It appears as if no deflection feedback was really used, but rather the piezo was kept at constant z. It seems the system is either drifting, or the cell is changing its mechanical properties (softening) or its height (shortening) due to the Ca2+ release. Could the authors comment on that? The fit obtained from fig 4d could be considered to have a slope of 1 (0.98), indicating an on-off response. More interestingly, the offset of 0.6s indicates that no matter the duration of the stimulation (and thus the total Ca2+ released), the time required to recover baseline levels is constant. Would that be an expected behaviour, or is this the case only because the released Ca2+ levels are small and very localized compared to the whole volume cell? In the discussion, the authors suggest that 'intracellular Ca2+ acts as an integrating signal that increases until it reaches the threshold Ca2+ concentration necessary to induce global Ca2+ response'. If that is the case, have they observed any global Ca2+ responses when using multiple cyclic indentations with very short recovery times? Taking into account their results on Ca2+ released for 1 single indentation, could they predict the number (or duration) of small indentations required to integrate enough Ca2+ signal to start a global Ca2+ response?", "responses": [] }, { "id": "5751", "date": "03 Sep 2014", "name": "Daniel Isabey", "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 by Lopez-Ayon et al. compares the calcium (Ca2+) response of osteoblasts to mechanical forces resulting from 2 types of well-defined micro-damages, namely a local membrane deformation and a micro-injury. An atomic force microscope is advantageously used to control and modulate these 2 types of stretching effects. In its present form, the article is already well structured, rich in new experimental results and easy to read. However, their message would be more complete if authors were considering the following points. Major comments Concerning methodological aspects:Since the probe calibration is very important for AFM measurements, authors should provide more details concerning this indispensable step. For instance, the method used to determine the spring constant of the probe is never mentioned. There is a non negligible hydro-drag force applied to the cantilever when it is moved at speed of 20 µm/s. If this aspect is not taken into account, the effective forces applied to cells and the energy spent to deform the membrane may be overestimated. A comment should be provided concerning this point. The method used to quantify the Ca2+ concentration is based on the analysis of averaged intensity in a predetermined zone drawn on the images acquired by fluorescent microscopy. So, signal intensity could depend on the layer which is in focus. The increase in intensity might thus be to the fact that cell is locally compressed leading to the generation of a calcium flow toward the basal cell face which is in focus. This additional effect should be appropriately discussed. Concerning Results and Discussion:The authors write in page 4 that: \"When consecutive low load indentation were performed at the same location on the same cell multiple local responses were induced\". This results is indeed comforted by the calcium concentration images shown at 6 s in Fig 2B. Multiple calcium spots are visible at 6 s on the non magnified cellular image, suggesting that a local mechanical effect may induce multiple (and durable) distant yet local effects. Such a mechanically induced signaling at a distance has been largely described and commented by the group of Ning Wang presently working at University of Illinois at Urbana Champaign (e.g., Wang et al., 2009). Clearly a tensed cytoskeleton plays a key role in such a distant phenomena and using, e.g., cytochalasin D, would have been of interest to prove that the distant local effects are related to the tensed cytoskeleton. Authors should at least comment on this interesting result revealed by the local membrane deformation test. Such an distant effect is not so evident in case of membrane rupture suggesting (since mechanical effect become global) important modifications for the cytoskeleton between the two types of stress studied. These distant effects could advantageously be analyzed and quantified. It is not clear whether the cells are tested in isolated conditions (like in Fig 2B) or with some degree of confluence (like in Fig 2E)... In the latter case, mechano-chemical interactions between cells could play a role in calcium signaling. This point should be clarified all along the paper. Detailed commentsFigures 1A, 1B seem to show the culture at higher confluence than 50-70% as used for experiments. The linear fits in Fig. 5 A, B, C don't represent the decay of peaks.", "responses": [] } ]	1	https://f1000research.com/articles/3-162
https://f1000research.com/articles/3-159/v1	11 Jul 14	{ "type": "Case Report", "title": "Case Report: Expanding the tumour spectrum associated with the Birt-Hogg-Dubé cancer susceptibility syndrome", "authors": [ "Patrick R. Benusiglio", "Sophie Gad", "Christophe Massard", "Edith Carton", "Elisabeth Longchampt", "Tiffany Faudot", "Jérôme Lamoril", "Sophie Ferlicot", "Sophie Gad", "Christophe Massard", "Edith Carton", "Elisabeth Longchampt", "Tiffany Faudot", "Jérôme Lamoril", "Sophie Ferlicot" ], "abstract": "Patients with the Birt-Hogg-Dubé cancer susceptibility syndrome are at high risk of developing renal cell carcinoma, pulmonary cysts and pneumothorax, and skin lesions called fibrofolliculomas. Here we report the case of a Birt-Hogg-Dubé patient with a primary clear cell carcinoma of the thyroid (a very rare type of thyroid cancer), and FLCN loss of heterozygosity within the tumour, providing molecular evidence for this association. Our findings expand the tumour spectrum associated with this syndrome. It is paramount to identify individuals with Birt-Hogg-Dubé so that they, and subsequently their affected relatives, can benefit from tailored cancer screening and prevention.", "keywords": [ "Birt-Hogg-Dubé", "clear cell carcinoma", "fibrofolliculoma", "hereditary cancer", "pneumothorax", "thyroid carcinoma." ], "content": "\n\nBirt-Hogg-Dubé (BHD) is a cancer susceptibility syndrome caused by dominantly-inherited mutations in the folliculin gene FLCN. Affected individuals are at risk of renal cell carcinoma (RCC), spontaneous pneumothorax associated with lung cysts and white skin papules called fibrofolliculomas1. RCC affects 34% of mutation carriers2, and most tumours are of chromophobe, oncocytoma, hybrid or clear cell histology3. Thirty-eight and 84% of BHD cases have a history of pneumothorax and fibrofolliculomas, respectively2. We report the case of a BHD patient with a primary clear cell carcinoma of the thyroid and provide molecular evidence supporting an association between the BHD syndrome and this rare tumour.\n\n\nCase report\n\nA 72 year-old French Caucasian male with a history of hypertension and early-stage rectal carcinoma (diagnosed at the age of 64) complained of a right thyroid nodule. Thyroid function tests were normal. Medication at the time consisted of irbesartan, lercanidipine and aspirin. The nodule was suspect on echography, and an operation was scheduled. Intraoperative fresh frozen analysis confirmed the malignant nature of the lesion, and thyroidectomy with paratracheal lymph node dissection was performed. The pathologist described a poorly circumscribed tumour measuring approximately 3.5 × 2 cm. On microscopic examination, nests of carcinomatous elements separated by a fibrous vascularized stroma were observed (Figure 1, hematoxylin and eosin). The neoplastic cells had a small, peripheral nucleus, with abundant clear cytoplasm. There was limited invasion of the capillaries, but widespread infiltration of the surrounding normal thyroid parenchyma and of the surgical margins. Staining was positive for cytokeratin 7 (Dako OV-TL 12/30), and negative for cytokeratin 20 (Dako Ks 20.8) and thyroglobulin (Dako DAK-Tg6). Lymph nodes were free of tumours. Histological examination was suggestive of a clear cell RCC metastatic to the thyroid, but no primary lesion was seen on 18-F fluorodeoxyglucose (FDG) positron emission tomography/computerised tomography (PET/CT) scans, in the kidneys or elsewhere. It was therefore concluded that the patient had a primary clear cell carcinoma of the thyroid. In the absence of papillary structures, significant nuclear grooves or pseudoinclusions, the tumour was considered a variant form of follicular carcinoma. Adjuvant radiotherapy was administered (46 Gy over five weeks). Two years later, multiple bilateral pulmonary nodules were seen on follow-up PET/CT scan. Two were surgically resected, and their microscopic appearance matched what had been observed in the thyroid two years earlier, except that complementary analyses now showed positive nuclear staining for Thyroid Transcription Factor 1 (TTF-1) (Figure 2), confirming as a result that the organ primarily affected was indeed the thyroid. There were no metastases in other organs, and the kidneys, as seen previously, were free of tumour.\n\nThe neoplastic cells have a small, peripheral nucleus, with abundant clear cytoplasm. Nests of carcinomatous elements are separated by a fibrous vascularized stroma. Hematoxylin and eosin staining, 5×.\n\nNuclear staining for TTF1 is strongly positive, 20×.\n\nThe patient also had multiple pulmonary air-filled cysts on baseline and follow up PET/CTs, as well as a right recurring pneumothorax. Family history was relevant as his son and two nephews had a history of spontaneous pneumothorax. On dermatological examination, one could see face fibrofolliculomas. Both these pulmonary and dermatological features were highly suggestive of BHD, and a blood sample was sent for FLCN analysis. Sequencing of the exons and of their flanking regions was performed with the Big Dye Terminator v.1.1 kit on the ABI 3730 sequencer (Applied Biosystems), and a search for large deletions was done using Multiplex Ligation-dependent Probe Amplification (MLPA, MRC-Holland). The c.1062G>C mutation in exon 9, which is classified as pathogenic by the SIFT (http://sift.jcvi.org/), Polyphen (http://genetics.bwh.harvard.edu/pph2/) and SNPs3D (http://www.snps3d.org/) bioinformatics prediction tools, was identified. It likely interferes with intron 10 splicing as it is located on the last base of exon 9, might affect a key splice site (http://www.umd.be/HSF/, http://genes.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html), and is adjacent to the already known c.1062+1G>A splicing mutation (https://grenada.lumc.nl/LOVD2/shared1/home.php). The diagnosis of BHD was therefore confirmed.\n\nTo investigate the association between BHD and the thyroid carcinoma, we performed FLCN analysis on tumoral and adjacent normal tissues. DNA was extracted from frozen sections with the QIAamp DNA mini kit (Qiagen), and exons and exon-intron junctions were sequenced with the methods decribed above. The c.1062G>C wild type allele was lost in the tumour, strongly suggesting loss of heterozygosity (LOH) (Figure 3).\n\nThe c.1062G>C wild type allele in exon 9 is lost in the tumour, while the patient is heterozygote for the mutation in the germline. Top sequence, control DNA. Middle sequence, germline DNA (normal tissue). Bottom sequence, tumour DNA.\n\nTwo years after the identification of the pulmonary metastases, the patient remains clinically well. His WHO performance status is 1. He has been included in a clinical trial at the Gustave Roussy Cancer Campus, and is now on temsirolimus and cetuximab. Previous lines of treatment with gemcitabine-oxaliplatine and docetaxel have had little effect on the tumour.\n\n\nDiscussion\n\nPrimary clear cell carcinoma of the thyroid is very rare. In two retrospective studies previously published, only three and four of 2784 and 572 thyroidectomies respectively were primary clear cell carcinomas4,5. Such a diagnosis is made when at least 75% of the tumour cells show marked cytoplasmic clearing6. This morphological pattern can occur in nearly all major thyroid tumour types, and is observed with the accumulation of vesicles derived from mitochondria, glycogen, lipid droplets, thyroglobulin or mucin7.\n\nTo our knowledge, this is the first time that the association between a non-renal clear cell carcinoma and BHD has been demonstrated. FLCN is a tumour suppressor gene, and associated tumours arise when both copies are inactivated most often through “second hit” somatic mutations8. In our patient, LOH provides molecular evidence for the inactivation of the second copy of FLCN. Interestingly, the microscopic features of thyroid clear cell carcinoma are similar to those of clear cell RCC, a tumour typically associated with BHD. The TTF-1 positivity and the absence of renal lesions on successive PET/CTs confirmed that our patient’s carcinoma originated in the thyroid, and not in the kidneys. As for the negativity for thyroglobulin, it was not unexpected since this staining is notoriously inconsistent in clear cell tumours of the thyroid9.\n\nOne should enquire about a personal or family history of BHD manifestations in patients with a diagnosis of thyroid clear cell carcinoma, and refer them for germline FLCN analysis when appropriate. Adult relatives of mutation carriers can then undergo targeted genetic testing. It is paramount to identify patients with BHD, as they are offered regular cancer screening with annual renal imaging (we alternate MRI and ultrasound imaging) and benefit from lifestyle recommendations. We advise patients not to smoke in order to minimize the risk of pneumothorax, and we inform them that activities such as deep sea diving or flying can trigger rupture of pulmonary cysts via changes in the atmospheric pressure, and that shortness of breath or chest pain in this context is likely due to pneumothorax10.\n\nOur report is of high scientific interest as, to our knowledge, no such case has ever been reported. In addition, we believe that it will increase awareness of BHD in the medical community, as this syndrome is too often overlooked even when obvious clinical manifestations are present.\n\n\nConsent\n\nThe patient described in this manuscript has provided informed written consent for his medical history and clinical images to appear in a scientific article.", "appendix": "Author contributions\n\n\n\nPRB, CM, and EC were involved in the clinical management of the case. PRB also provided genetic counselling and wrote the manuscript (with support from the other authors). SG and TF performed the genetic analyses on the tumour. JL performed germline genetic analyses on the DNA extracted from blood. EL and SF were involved as pathologists. 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\nFinancial support to the PREDIR Centre (PRB, SF) and the INSERM U753 Research Unit (SG) is provided continuously by the Ligue Nationale contre le Cancer (Comités de l’Indre et du Cher), the Myrovlytis Trust (BHD Foundation) and the Fondation Gustave Roussy.\n\n\nAcknowledgments\n\nWe thank Mrs Frederique Legale who provided valuable secretarial assistance.\n\n\nReferences\n\nMenko FH, van Steensel MA, Giraud S, et al.: European BHD Consortium. Birt-Hogg-Dube syndrome: diagnosis and management. Lancet Oncol. 2009; 10(12): 1199–206. PubMed Abstract \| Publisher Full Text\n\nToro JR, Wei MH, Glenn GM, et al.: BHD mutations, clinical and molecular genetic investigations of Birt-Hogg-Dube syndrome: a new series of 50 families and a review of published reports. J Med Genet. 2008; 45(6): 321–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPavlovich CP, Walther MM, Eyler RA, et al.: Renal tumors in the Birt-Hogg-Dube syndrome. Am J Surg Pathol. 2002; 26(12): 1542–52. PubMed Abstract \| Publisher Full Text\n\nShimizu K, Nagahama M, Kitamura Y, et al.: Clinicopathological study of clear-cell tumors of the thyroid: an evaluation of 22 cases. Surg Today. 1995; 25(12): 1015–22. PubMed Abstract \| Publisher Full Text\n\nYang GC, Fried K, Scognamiglio T: Cytological features of clear cell thyroid tumors, including a papillary thyroid carcinoma with prominent hobnail features. Diagn Cytopathol. 2013; 41(19): 757–61. PubMed Abstract \| Publisher Full Text\n\nRosai J, Carcangiu ML, DeLelli RA: Atlas of Tumor Pathology: Tumors of the thyroid gland. Armed Forces Institute of Pathology. 1992. Reference Source\n\nCarcangiu ML, Sibley RK, Rosai J: Clear cell change in primary thyroid tumors. A study of 38 cases. Am J Surg Pathol. 1985; 9(10): 705–22. PubMed Abstract \| Publisher Full Text\n\nVocke CD, Yang Y, Pavlovich CP, et al.: High frequency of somatic frameshift BHD gene mutations in Birt-Hogg-Dubé-associated renal tumors. J Natl Cancer Inst. 2005; 97(12): 931–5. PubMed Abstract \| Publisher Full Text\n\nRopp BG, Solomides C, Palazzo J, et al.: Follicular carcinoma of the thyroid with extensive clear-cell differentiation: a potential diagnostic pitfall. Diagn Cytopathol. 2000; 23(3): 222–3. PubMed Abstract \| Publisher Full Text\n\nGupta N, Seyama K, McCormack FX: Pulmonary manifestations of Birt-Hogg-Dubé syndrome. Fam Cancer. 2013; 12(3): 387–96. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5429", "date": "22 Aug 2014", "name": "Nicole Philip", "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 short report provides new knowledge on a very rare disorder, the Birt-Hogg-Dube syndrome. The presence of LOH in the tumor is a strong argument to conclude that the tumor is related to the syndrome.This new information can be useful for clinicians to diagnose new patients and offer appropriate surveillance..", "responses": [] }, { "id": "6020", "date": "03 Sep 2014", "name": "Andre-Pascal Sappino", "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 case report submitted by Benusiglio et al. is well documented and provides convincing evidence that we should consider widening the tumor spectrum generally ascribed to the rare Birt-Hogg-Dubé syndrome. Indeed, the authors report the case of a 72 y-old patient with a suggestive familial and personal history who developed a primary clear-cell carcinoma of the thyroid. Gene-testing for FLCN revelead a germ-line deleterious mutation and LOH for FLCN was demonstrated in the thyroid tumoral tissue. This finding is of interest since most (all ?) cancers described so far in this syndrome arise from the kidneys.Altogether, the title, the abstract, and the article content, based on adequate methodology, are appropriate.However, we are told that the reported patient developed at the age of 64 an early-stage rectal cancer. It would have been interesting to know more about the morphological aspects of this primary digestive tumor and if the authors could also perform on this primary the search for FLCN LOH. This information may be relevant since the median age of tumor diagnosis in this syndrome is generally considered to occur below 50 y, a point that the authors fail to discuss.", "responses": [] } ]	1	https://f1000research.com/articles/3-159
https://f1000research.com/articles/3-158/v1	10 Jul 14	{ "type": "Research Article", "title": "Comparison of phenothrin mousse, phenothrin lotion, and wet-combing for treatment of head louse infestation in the UK: a pragmatic randomised, controlled, assessor blind trial", "authors": [ "Ian F. Burgess" ], "abstract": "In this investigation of effectiveness of an alternative pediculicide dosage form, we recruited 228 children and 50 adult participants from Bedfordshire, UK, to a randomised, controlled, assessor blind trial comparing two insecticide products with mechanical removal of lice as a control group. Participants using insecticide were treated with either the investigative 0.5% phenothrin mousse, for 30 minutes, or 0.2% phenothrin lotion, for 2 hours as the reference product. Both treatments were applied only once, followed by shampoo washing. Those treated by wet-combing with conditioner were combed 4 times over 12 days. Parents/carers carried out the treatments to mimic normal consumer use. The outcome measure was the absence of lice, 14 days after treatment for the insecticides, and up to 14 days after completion of combing. Intention to treat analysis of the outcomes for 275 participants showed success for phenothrin mousse in 21/105 (20.0%), in 23/107 (21.5%) for phenothrin lotion, and in 12/63 (19.1%) for wet-combing. People receiving mousse were 1.07 (95% CI, 0.63 to 1.81) times more likely to still have lice after treatment compared with those treated with lotion. The group of participants who received the wet combing treatment were 1.13 (95% CI, 0.61 to 2.11) times more likely to still have lice after the treatment. None of the treatments was significantly (p < 0.05) more effective than any other. This study was carried out in an area where moderate resistance to phenothrin was demonstrated after the study by using a bioassay. Analysis of post treatment assessments found that failure of insecticides to kill louse eggs had influenced the outcome.", "keywords": [ "test" ], "content": "Introduction\n\nDuring the early 1990s, a number of synthetic pyrethroid-based formulations for treating head louse infestation were introduced into the British market. The majority of these used d-phenothrin as the active substance in a variety of dosage forms1. In 1997 a phenothrin mousse was developed based on a concept developed in Australia using natural pyrethrum2. The aim of the new product was designed to be more manageable during application and thus gain greater consumer acceptability than existing preparations. Laboratory studies of phenothrin lotion had indicated a high level of activity for the insecticide, so theoretically it could be incorporated into a formulation requiring a shorter application time1. However, at this time there was increasing evidence of insecticide resistance in several areas of the UK3,4, in parallel with a renewed consumer interest for treating head louse infestation by combing, either as the principal measure or as a component of conventional insecticide treatment. At the time of the study the most widely promoted combing method was “Bug-Busting” (Community Hygiene Concern, London), which used a fine toothed plastic comb for wet combing with conditioner, repeated at 3–4 day intervals for 2 weeks. It suggested the first combing could remove all lice so only newly hatched nymphs would be found during subsequent combing sessions before they could mature and lay eggs5. Before our investigation only two studies of wet-combing had been conducted. In one Bug-Busting was half as effective as two applications of malathion lotion6. A second found it more effective than permethrin creme rinse, but the dropout rate from both treatments made the interpretation of the results difficult7.\n\nWe performed a pragmatic, observer blinded, three armed clinical trial analysed by intention to treat, comparing single applications of 0.5% d-phenothrin mousse or 0.2% d-phenothrin lotion with the Bug-Busting protocol of wet-combing with conditioner. The study was designed to evaluate the effectiveness of each treatment when in use by the public.\n\n\nMethods\n\nWe recruited participants (children and adults) to the study from respondents to an information letter distributed through schools or via general practitioners associated with Bedfordshire Health. Prospective participants or their parents/carers telephoned the study co-ordinator to make an appointment for a home visit by a trained agency nurse. Most visits were within 24 hours unless requested at a different time. Nurses followed a standard approach to check for living lice using a plastic detection comb (Albyn of Stonehaven Ltd, Stonehaven, Scotland). If moving head lice were found, and the individual was 4 years or over, they were invited to join the study. Prospective participants were conducted through a standard consent procedure in which the content of information sheet was explained verbally and the parent/carer confirmed that they understood the function, processes, and commitments of the study before signing the consent form, which was witnessed by an independent adult. Participants were individually assigned a randomised treatment. All other household members were offered examination and, if found to be infested with head lice, were given the opportunity to join the study.\n\nPrior to inclusion in the study, all participants provided baseline data including: age, gender, ethnicity, hair characteristics including length, thickness, degree of curl, and previous pediculicide use. Inclusion criteria required availability for up to 28 days to accommodate each of the possible treatment regimens and a suitable adult available to perform or assist in application of the treatment. Candidates excluded from this study were: pregnant or nursing mothers; anyone who had bleached, colour-treated or permanently waved their hair; used pediculicide, had been treated with antibiotics, or had participated in a clinical trial during the 4 weeks prior to this trial. In addition anyone with sensitivity to any pyrethroid insecticide or chrysanthemums; receiving treatment for asthma; suffering from a persistent skin disorder of the scalp (other than head lice); or had already participated in this study; was also excluded.\n\nHousehold members with lice who either did not wish to participate or who failed to satisfy the inclusion criteria were given advice about appropriate treatment methods. Any participants found with lice after completion of the study period were supplied with 0.5% malathion lotion (Suleo-M lotion, Seton Healthcare Group plc, Oldham, UK) in conformity with the Bedfordshire Health policy for treatment. No payment was offered for participation.\n\nEthical approval was granted by both the South Bedfordshire and North Bedfordshire Local Research Ethics Committees of Bedfordshire Health, one of the conditions of which was that the data would be published in the public domain. The study was registered with the Current Controlled Trials database (ISRCTN73201839). See Supplementary files for the study protocol.\n\nThe study was conducted in compliance with Good Clinical Practices, and in conformity with the principles of the Declaration of Helsinki and European Standard, EN540: Clinical investigation of medical devices for human subjects. Written and witnessed informed consent was obtained from the participants and parents or guardians of children under 18 years of age.\n\nThe nurses explained how to apply the treatment to the parent or carer and also gave a printed copy of these instructions. The parent then applied the treatment in the presence of the investigator. The nurses were instructed to answer any questions, and to note these, but not to intervene if any error was observed during treatment. This was intended to represent the clinical situation.\n\nWe supplied one group of participants with 0.5% d-phenothrin mousse in 50ml butane pressurised containers, with a canula fitting to allow direction of the mousse during delivery (Full Marks Mousse, Seton Healthcare Group plc, Oldham, UK). This also contained citrate buffered water, ethanol, and emulsifying wax. We supplied the second group with 0.2% d-phenothrin water/isopropanol lotion in 50ml glass bottles with a dropper aperture (Full Marks Lotion, Seton Healthcare Group plc). Carers applied the products to dry hair to saturate the hair and scalp. We made available as many containers of product necessary to comply with the instructions. People using mousse washed it from the hair with shampoo after 30 minutes. Those treated with the lotion left it on the hair for 2 hours before shampooing. We supplied both groups with a non-medicated shampoo (L’Oréal Children’s Shampoo, L’Oréal (UK) Ltd, London, UK) for this. Both groups received a single application of treatment.\n\nWe supplied the third group of participants with a “Bug Buster” pack (Community Hygiene Concern, London, UK) for performing the wet-combing technique. We also supplied a bottle of the same non-medicated toiletry shampoo and four 60ml bottles of conditioner rinse (one for each treatment day) (L’Oréal Children’s Shampoo and L’Oreal Conditioner, L’Oréal (UK) Ltd,), and a diary card. Parents/carers were instructed to wash the participant’s hair with the shampoo, massage in a generous amount of conditioner, and comb through the hair systematically from scalp to tip with the louse removal comb provided in the pack. They wiped the comb on a paper towel between strokes. Combing was repeated at regular intervals for 2 weeks (days 0, 4, 8, 12 of the study). We asked carers to fix any lice found during the combing onto the diary card using clear cellulose adhesive tape and to record how much time was spent combing.\n\nWe made mousse and lotion treatment follow ups on days 4, 7, 10, and 14 and those for wet-combing on days 14, 21, and 28 after commencement of treatment. Day 14 was used as the point for measure of primary outcome for the insecticide groups. However, in the wet-combing group it was possible there could be viable eggs present on day 14 and therefore follow up examinations were also conducted on days 21 and 28 to detect any emerging nymphs or other lice missed at day 14. During each of the follow up assessment examinations, any lice found by detection combing were removed and fixed to the case record using clear cellulose adhesive tape. These were later examined in the laboratory to determine the gender or development stage of each insect.\n\nNurses collected the containers of mousse and lotion after treatment so that the quantities used could be measured. Bottles of conditioning rinse from the wet combing group were also collected for measurement after the final assessment on day 28, although some of these were mislaid by participants.\n\nWe estimated sample sizes to show a difference between wet combing and phenothrin lotion treatment with 95% confidence, 90% power, and equivalence between the two phenothrin groups to within 20%. For this calculation we conservatively estimated that the phenothrin products would exhibit approximately 80% effectiveness and wet combing 50% success. Sample size calculations were made by the sponsor’s consultant statistician who estimated a minimum sample size of 104 participants in each phenothrin treated group and 58 participants treated with wet combing (266 evaluable participants) would satisfy this probability with greater than 90% power.\n\nA computer generated list, prepared by the sponsor’s statistician, was used for randomisation of treatments, made up of balanced blocks of 133 treatment allocations with a relative frequency for each of the treatments of 52:52:29. Randomisation was by individual so that different members of a household could receive different treatments. Nurses involved in recruitment were supplied with envelopes in batches of ten and asked to issue them sequentially. Investigators, who were unaware of which treatment had been used, made follow up examinations using plastic detection combs to check for the presence of living lice. On day 4, 7, and 10 these assessors were a different group of agency nurses. On days 14, 21, and 28 the assessing investigators were from the Medical Entomology Centre (IFB and CMB).\n\nWe analysed for differences between groups based on the intention to treat (ITT) population and tested equivalence only using the per-protocol (PP) population. We calculated differences in cure rate using a chi-squared test and equivalence to within 20% based on the 95% confidence limits derived from the normal approximation to the binomial distribution. The initial analyses for the sponsor were performed by the consultant statistician (PN Lee Statistics and Computing Ltd, Sutton, UK) using bespoke software. Post hoc analyses performed by the investigators employed Epi-Info version 6, OXSTAT II version 1.11, and purpose built spreadsheet calculators. Differences between groups in baseline characteristics, safety, acceptability, and efficacy were tested using Fisher’s exact test for yes/no variables and the Mann-Whitney U test for ranked variables.\n\nDuring the course of the study a high level of treatment failure was observed in the insecticide groups. We collected samples of lice at the final assessment from five participants from different parts of the study area for a bioassay evaluation for sensitivity to phenothrin. The insects were placed on treated or control filter papers using a method previously described for tests of permethrin sensitivity3. Each treated filter disc was impregnated with 500 µl of 2% d-phenothrin solution, giving an insecticide deposition rate of 157 μg cm-2. The mortality outcomes of the tests were compared, with a baseline sensitivity obtained using laboratory-reared, insecticide sensitive, body/clothing lice by means of log-probit analysis using LDP Line software.\n\n\nResults\n\nThe study was conducted between June 1997 and March 1998, during which informed consent was obtained for 228 children and 50 adults to participate (Figure 1). Two people were excluded from further analysis as, upon inspection, no live lice had been found. The recruitment case record form for one other participant was lost so this case was also excluded from the study.\n\nThe 275 participants were randomly assigned to one of the treatment groups: 105 received phenothrin mousse, 107 phenothrin lotion, and 63 were allocated to the wet-combing treatment. Of these participants, 246 (89.5%) (100 mousse, 100 lotion, and 46 wet-combing) completed the trial with adequately complete follow-up data sets (Figure 1 and Dataset 1). From the original study group, 5 participants (2 treated with lotion and 3 wet-combing) were non-compliant and excluded from the per-protocol analyses. Non-compliance involved additional combing or other unauthorised treatments (2 receiving lotion and 2 wet-combing) and one participant on wet-combing shaved his head. There were 24 other withdrawals: 2 people on wet-combing dropped out; 20 were lost to follow-up (5 from the mousse group, 5 from lotion, and 10 from wet-combing); and 2 people from the wet-combing group were not allocated Day 21 or Day 28 appointments in error following a communication failure between the study coordinators and the investigators conducting final assessments. Two of those lost to follow up were due to bereavement and the two drop outs chose not to continue in the study. The rate of protocol violation/withdrawal was significantly higher (p < 0.001) in the wet-combing group than the phenothrin-treated groups.\n\nOf the 275 people known to satisfy the inclusion and exclusion criteria, 62 (23%) were male (Table 1). The percentage distribution of males was similar for the two phenothrin groups, 18% and 21% respectively, but higher in the wet-comb group (33%). This difference in proportion of males between wet-comb group and mousse-treated group was statistically significant (p < 0.05) and the difference between wet-comb group and the lotion-treated group was nearly significant (0.05 < p < 0.1).\n\nLevels of statistical variation between groups: Figures in bold type show the group exhibiting a statistical disparity indicating possible randomisation anomalies.\n\n* Difference significant at p < 0.1; ** Difference significant at p < 0.05; § Difference significant at p < 0.01\n\nPost-treatment examinations at day 14 showed that there were 20/107 successful treatments and 3 cases of reinfestation after cure (an overall success rate of 21.5%) using phenothrin lotion and 18/105 successes and 3 cases of reinfestation for phenothrin mousse (giving 20.0% overall success). This made mousse users 1.07 times more likely to have lice after completion of the treatment (95% confidence interval (CI) 0.63 to 1.81; odds ratio (OR) 1.10, 95% CI 0.56 to 2.13). In the case of wet-combing with conditioner there were 12/63 (19.1%) successful treatments and no cases of reinfestation. Participants treated with combing were, therefore, 1.13 times more likely to have lice (95% CI 0.61 to 2.11; OR 1.16, 95% CI 0.53 to 2.54) than if they had been treated with phenothrin lotion. People treated with wet-combing were also 1.05 times more likely to have lice than those participants receiving phenothrin mousse (95% CI 0.56 to 1.99; OR 1.06, 95% CI 0.48 to 2.34).\n\nIn both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 first were stage nymphs that could only have originated from eggs not killed by insecticide. However, it was not possible to properly analyse the full effect of ovicidal failure due to participants being withdrawn early in the study on grounds of lack of efficacy. Nevertheless from the data available it was possible to determine that failure to kill louse eggs was a major contributing factor in the low rate of outcome success.\n\nIntention to treat analysis found no statistical difference (p < 0.05) between the lotion and mousse. Similarly, no statistical difference was found between wet-combing and either of the insecticides. Success in curing the infestation was also not significantly associated with gender or hair type, thickness, or length, after adjustment for any randomisation anomalies. However, success rates were significantly (p < 0.01) higher in people who had previously used a head louse treatment successfully, by an estimated factor of 1.88 (95% CI 1.13 to 3.11). Success also significantly (p < 0.01) increased with age.\n\nAnalysis based on the per-protocol population, without taking into account the failure of randomisation, showed no significant difference in success rates between the two phenothrin groups (lotion 17.8%, mousse 13.7%), which showed equivalence to within 20% (mousse-lotion difference, -4.1%, 95% CI -6.1% to 14.4%).\n\nWe found several clearly defined treatment-related adverse events in people treated using the phenothrin products: 12 adverse events in 11 people using lotion (9 scalp irritation, 3 irritation of the respiratory system); 10 adverse events in six people treated with mousse (5 scalp irritation, 3 dry skin, 1 bullous reaction, 1 paraesthesia of the scalp). There were no similarly defined adverse events for the wet-combing group but five carers reported children expressing signs of stress while being combed, one person reported discomfort during combing, and backache or arm/shoulder aches for three carers were also reported, but not formalised as reported adverse events. All events were considered mild and resolved rapidly except for two cases. The case of paraesthesia, which was classified as moderate, persisted for two days after treatment, and one case of dry skin persisted for some time after treatment but may have been an exacerbation of a pre-existing problem. Stinging of the hands and paraesthesia-like reactions were also reported by some of the carers while applying the phenothrin-based products. Paraesthesia has been reported from use of other pyrethroid preparations8–10 and would likely be exacerbated by the presence of alcohol in the product.\n\nLice from different participants showed marked differences when tested for sensitivity to phenothrin (Table 2). In all cases the lice were taken from people who had experienced treatment failure during the study so it was not surprising that the majority had insects that were resistant to the insecticide. However, one person had apparently also been reinfested with sensitive lice from a contact, as shown by the mixed sensitivity of the insects. All lice from another participant, treated using wet-combing, were phenothrin susceptible. Output data from the LDP Line analyses of the observations of head lice, in comparison with susceptible laboratory reared body lice, showed a resistance ratio (RR) of 54.74 when exposed to the insecticide. From the log-probit analyses the estimated time required to kill 50% of the insects (LT50) was 502 minutes based on a mortality curve with a slope of 1.0096 ± 0.1324 (chi-squared 19.9681, p = 0.0005). In contrast the estimated LT50 for the body lice was 24.74 minutes (slope = 5.1932 ± 0.5086, chi-squared = 11.6217, p = 0.0404). Log-probit analysis also suggested that the insecticide sensitive head lice (LT50, 44 minutes; LT95, 95 minutes) were approximately twice as tolerant of phenothrin as laboratory reared lice, although the number of insects involved was too small to provide a clear distinction.\n\n* The time of death of these lice was sufficiently delayed that it may have been due to dehydration.\n\n** Lice survived for longer than 540 minutes when provided with a blood meal.\n\n\nDiscussion\n\nThis was a pragmatic study designed to investigate the three treatments under conditions that mimicked normal use by parents and carers. The outcomes demonstrated that all three products were less than adequate to eliminate head louse infestation when used according to instructions.\n\nIn the case of the phenothrin mousse, a major limitation of the product in use was that the foam was too dry and insufficiently thermo-labile. It did not break down to a fluid that could be readily spread through the hair. This problem probably arose because too little alcohol was included in the formulation so the foam generated from water and emulsifying wax did not break physically after application. Consequently, many of the parent carers probably failed to achieve an adequate or even coverage of the hair and scalp when applying it. In contrast, phenothrin lotion was too fluid, like other alcohol based products. This meant that in use it was easy to apply too little product because a small volume of the fluid made the hair look wet and, by implication, thoroughly coated. Also, this formulation did not contain any of the terpenes that had been shown to contribute much of the ovicidal activity shown by other alcoholic lotions11. At that time the “Bug Buster” pack contained a two-part comb that was initially described as “unique, safe and well researched” and, “reliable even though its use may be time consuming”12. Although when the product was shown to be relatively ineffective by independent investigators6 they were criticised by the pack suppliers for using a product that had been superseded by the time of publication13. Our investigators found it relatively easy to find lice on heads that parents using the “Bug Buster” comb believed were louse free. Therefore, it can be concluded that the two-part comb (Figure 2) was not as effective as originally claimed and was probably not as easy to use as either the plastic detection comb we used or its replacement, which was similar to our detection comb.\n\nFor this study the sponsor engaged a general nursing agency to supply staff to perform the majority of study functions in place of trained investigators as a cost saving measure. We believe this was not only a false economy in terms of data quality but may also have resulted in a breakdown of Good Clinical Practice. We could not determine whether some of the nurses failed to understand their responsibilities as investigators, or simply could not follow instructions, but some of the apparent failures of randomisation could be attributed to poor practice. For example, people in the wet-comb group were significantly more likely to have short hair (p < 0.05) or fine hair (p < 0.05) than those in the mousse-treated group. These anomalies were attributed to a failure of randomisation at the point of enrolment by some nurses engaged in recruitment. Early in the study we drew this possibility the attention of the sponsor, as well as to the management of the nursing agency, who assured us that correct procedures were being followed. Nevertheless, a failure of randomisation was identified at analysis because we suspected a general reluctance on the part of carers and children to participate in the combing group. Boys with shorter hair were apparently seen as an easier option for wet-combing so it is possible the numbered envelopes were opened before they were specifically allocated to individuals in households were several members were participating. Therefore we adjusted for this anomaly using stratified chi-squared analyses, although ultimately it made little difference to the overall outcome analysis.\n\nWe detected another, less easily identified, anomaly in that people who had a previous experience of a successful treatment with an insecticide lotion were more likely to be recruited to the lotion group (lotion v mousse p < 0.01; lotion v wet combing p < 0.05) (Table 1), although there was no evidence this was due to a failure of randomisation at the point of allocation.\n\nAnother deviation arose because most participants that were found to have lice at post-treatment assessments on day 4 or later were not withdrawn and provided with rescue treatment by the agency nurses but allowed to remain in the study, in some cases, until assessed by us on day 14. As far as we were able to determine, this did not affect the outcomes, and was partly associated with the logistical difficulties of transporting large volumes of documentation between offices over a short period of time, but again suggests a lack of understanding of the requirements of the study.\n\nIn the study area the local policy had been to use malathion for head louse infestation during the previous few years, so we did not anticipate significant resistance to pyrethroids. However, the small sample of lice we collected during the later stages of the study demonstrated that resistance was present and probably existed over much of the study area. The level of treatment failure could not be fully explained by resistance because five years later we observed a 75% success rate using a phenothrin aqueous emulsion in a study area that overlapped with this geographically14. Rather the problem may have rested with the preparations used because another study using the alcoholic phenothrin lotion, conducted in a different part of the country, obtained a similarly poor result with only 2/15 (13.3%) cures15.\n\nThis study was conducted because the Medicines Control Agency (MCA), now the Medicines and Healthcare Products Regulatory Agency, did not consider clinical investigations of the mousse in India representative of conditions likely to be encountered in Britain, so the manufacturers were required to conduct a UK-based study for confirmation of efficacy. Surprisingly however, before we could complete the analysis of this study, the MCA issued a Marketing Authorisation for the phenothrin mousse. As a result the product was launched without further assessment by the MCA. Not surprisingly, given the poor effectiveness observed in this clinical investigation, there was widespread anecdotal reporting of treatment failure by consumers after using the product. Nevertheless it remained in the UK market until 2009.\n\nThere are lessons that can be learned from this experience. The first is that bioassay tests conducted in a laboratory, whether using laboratory reared insects or even wild collected ones, can only be indicative of efficacy for a formulation1,11 and it is unknown for a treatment to perform poorly in vitro yet be effective in vivo. Of course, laboratory reared lice and other ectoparasites are usually poorly representative of the physiological characteristics of those found on their natural hosts, especially with regard to characteristics such as resistance. Even ex vivo screens, using insects recently collected from the wild, may be only partially representative and several replicate tests should be performed using insects from geographically separated locations to ensure that the outcome is not obtained either by chance or due to some happenstance of physiological difference in the insects from that location. Attempting to draw any kind of conclusion about efficacy from only two or three lice or a single replicate test is fraught with risk, although this appears to have been a common practice in some investigations16–21.\n\nThe second lesson, one that was recognised by the Medicines Control Agency when they initially insisted that the phenothrin mousse should be clinically tested in the UK, is that a clinical investigation of a pediculicide in a country where treatments for head lice are not routinely used is not likely to be representative of the possible outcomes in the territory where the product is destined to be marketed. Such studies may be indicative of possible outcomes but basing strong claims about how a product will work in a developed country, where lice are regularly exposed to a range of chemical entities, on the results of studies conducted in a developing country is just as flawed as relying on in vitro data. However, in most developed countries there are products that appear to have been evaluated only in trials in developing countries2,22–25.\n\nThe third, and to us, the most significant point is that the efficacy data for pediculicides must be a high priority for regulatory authorities before granting a Marketing Authorisation (MA). This is significant because the products are used on children and must be safe, clinically effective, and also cost effective. In this instance the product continued to be marketed after the final data were available, despite clearly showing its lack of effectiveness. This raises a question of how many other products could be on the market without evidence of efficacy. No doubt makers of such products rely on consumer feedback and complaints as a guideline as to whether their products are both acceptable and effective. However, in practice, most manufacturers receive relatively few complaints about efficacy and few pediculicides have been subjected to the kind of post marketing surveillance applied to some other medications.\n\nEven in the rather more litigious circumstances prevailing in the United States of America, there have been few who have gone as far as legal action to press claims of inefficacy. In overall terms it was found that the legal complaint process was hindered by the regulatory process, as argued by the defence attorneys acting for various drug companies in one class action stating “..The claims (of the Plaintiffs) stand in direct conflict with the Food Drug and Cosmetic Act, moreover, because the “defendants’ medications cannot be sold for the treatment of head lice and labeled to say that the medications are not effective when simultaneously federal law and regulations require the labeling to say that the products are effective.”26 (http://www.law360.com/articles/36260/drug-makers-fight-class-action-over-lice-treatment) thereby turning the onus for verification of efficacy back to the competent authority, in this case the federal Food and Drug Administration (FDA), although in another plaint the Texas Supreme Court gave a per curiam ruling “...that the FDCA contains no such “complex and interrelated federal scheme of law, remedy, and administration” that would divest the state courts of jurisdiction..”27 (http://www.supreme.courts.state.tx.us/historical/2005/feb/031052.htm) suggesting that those courts could, if they so chose, declare products ineffective and presumably thereby place the FDA in an difficult position with respect to its approval of certain preparations27. Therefore, only products for which adequate clinical studies have been conducted, and then those data placed in the public domain with appropriate periodic review to ensure resistance has not affected efficacy, can be considered effective. Simply relying on a competent authority MA is not adequate justification for continuing to sell a product when there are doubts about its effectiveness, as we highlighted in presenting the results from one of our recent investigations28. Just because products or active substances may have been effective when first introduced29 does not mean that they remain so, as indicated by more recent clinical investigations using some so-called “standard of care” products as comparators30,31, and both industry and regulators should be responsive to changes in circumstance.\n\n\nData availability\n\nF1000Research: Dataset 1. Individual demographic data collected from participants and lice found by detection combing post-treatment, 10.5256/f1000research.2026.d3172832\n\n\nParticipant consent\n\nWritten and witnessed informed consent was obtained from the participants and parents or guardians of children under 18 years of age.", "appendix": "Author contributions\n\n\n\nAll authors contributed to the original study design. PN was responsible for the ethics applications and liaison with the community health authorities, and was the Supervising Clinician throughout the study. IFB and CMB contributed to the logistics of participant management, conducted all final assessments on Days 14 to 28, and evaluated the collections of lice on the Case Record Forms for determination of treatment outcome. IFB conducted resistance testing of lice, analysis of those results, and post hoc analyses on the historical data to correct some errors and omissions. IFB was responsible for drafting the manuscript and all three authors read the draft and commented on its contents.\n\n\nCompeting interests\n\n\n\nIFB is a consultant to several large, medium, and small companies that develop and market head louse treatment products, including medicinal products, medical devices, cosmeceuticals, and combs. CMB and PN are not aware of any competing interests.\n\n\nGrant information\n\nThis study was funded on a commercial basis by Seton Healthcare Group plc.\n\n\nAcknowledgements\n\nThe conduct of this study was supported financially by Seton Healthcare Group plc, which contributed to the original design of the study but played no role in the decision to publish the results, permission for which was withheld while the company was in existence as an independent commercial entity. All confidentiality agreements between the investigators and the sponsor in relation to this work have now expired. Reckitt Benckiser plc, the current beneficiary of the legacy Seton Healthcare assets for this category of product, has kindly granted approval for publication of the manuscript and the original study protocol. The study was supported logistically by Bedfordshire Health Authority and Bedfordshire Family Health Authority, Luton, both of which have now been replaced by other services with different responsibilities and jurisdictions during various reorganisations of the National Health Service in the UK. Our thanks go to Deirdre Power and Stanley Lindsay who were the Study co-ordinators on behalf of South Bedfordshire Community Healthcare Trust. All treatments were supervised and follow up assessments performed, except days 14 to 28, by staff from Independent Nursing Services Ltd, Wootton, Bedfordshire. Thanks also to all the schools and General Practitioners who circulated Information Sheets to families in their care, as prospective participants. The original statistical analyses were conducted by Peter Lee, PN Lee Statistics and Computing Ltd, Sutton, Surrey, on behalf of the sponsor, Seton Healthcare Group plc. Sara Gray and Barbara Temesi, both of Seton Healthcare Group plc, monitored the study documentation on behalf of the sponsor.\n\n\n\n\nSupplementary files: https://f1000researchdata.s3.amazonaws.com/supplementary/2026/0086c068-7ac2-42a1-8135-946436252380.zip\n\n\nReferences\n\nBurgess I, Veal L, Sindle T: The efficacy of d-phenothrin and permethrin formulations against head lice: a comparison. Pharm J. 1992; 249: 692–93. Reference Source\n\nBurgess IF, Brown CM, Burgess NA: Synergized pyrethrin mousse, a new approach to head lice eradication: efficacy in field and laboratory studies. Clin Ther. 1994; 16(1): 57–64. PubMed Abstract\n\nBurgess IF, Brown CM, Peock S, et al.: Head lice resistant to pyrethroid insecticides in Britain. BMJ. 1995; 311(7007): 752. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowns AMR, Stafford KA, Harvey I, et al.: Evidence for double resistance to permethrin and malathion in head lice. Br J Dermatol. 1999; 141(3): 508–511. PubMed Abstract \| Publisher Full Text\n\nIbarra J, Fry F, Wickenden C, et al.: Overcoming health inequalities by using the Bug Busting ‘whole-school approach’ to eradicate head lice. J Clin Nurs. 2007; 16(10): 1955–1965. PubMed Abstract \| Publisher Full Text\n\nRoberts RJ, Casey D, Morgan DA, et al.: Comparison of wet combing with malathion for treatment of head lice in the UK: a pragmatic randomised controlled trial. Lancet. 2000; 356(9229): 540–544. PubMed Abstract \| Publisher Full Text\n\nBingham P, Kirk S, Hill N, et al.: The methodology and operation of a pilot randomized control trial of the effectiveness of the Bug Busting method against a single application insecticide product for head louse treatment. Public Health. 2000; 114(4): 265–268. PubMed Abstract \| Publisher Full Text\n\nFlannigan SA, Tucker SB, Key MM, et al.: Synthetic pyrethroid insecticides: a dermatological evaluation. Br J Ind Med. 1985; 42(6): 363–372. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilks MF: Pyrethroid-induced paresthesia--a central or local toxic effect? J Toxicol Clin Toxicol. 2000; 38(2): 103–105. PubMed Abstract \| Publisher Full Text\n\nJiang N, Nutter TJ, Cooper BY: Molecular and cellular influences of permethrin on mammalian nociceptors at physiological temperatures. Neurotoxicology. 2013; 37: 207–219. PubMed Abstract \| Publisher Full Text\n\nBurgess I: Malathion lotions for head lice - a less reliable treatment than commonly believed. Pharm J. 1991; 247: 630–632. Reference Source\n\nIbarra J: In: Proceedings of the North Thames Head Lice Symposium “Controversies in head lice treatment”. North Thames Drug Information Centre, London. 1997; 13–17.\n\nIbarra J, Fry F, Wickenden C: Treatment of head lice. Lancet. 2000; 356(9246): 2007. PubMed Abstract \| Publisher Full Text\n\nBurgess IF, Brown CM, Lee PN: Treatment of head louse infestation with 4% dimeticone lotion: randomised controlled equivalence trial. BMJ. 2005; 330(7505): 1423–1425. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPlastow L, Luthra M, Powell R, et al.: Head lice infestation: bug busting vs. traditional treatment. J Clin Nurs. 2001; 10(6): 775–783. PubMed Abstract \| Publisher Full Text\n\nToloza AC, Zygadlo J, Mougabure-Cueto G, et al.: The fumigant and repellent activity of aliphatic lactones against Pediculus humanus capitis (Anoplura: Pediculidae). Mem Inst Oswaldo Cruz. 2006; 101(1): 55–56. PubMed Abstract \| Publisher Full Text\n\nHeukelbach J, Oliveira FA, Speare R: A new shampoo based on neem (Azadirachta indica) is highly effective against head lice in vitro. Parasitol Res. 2006; 99(4): 353–356. PubMed Abstract \| Publisher Full Text\n\nToloza AC, Zygadlo J, Cueto GM, et al.: Fumigant and repellent properties of essential oils and component compounds against permethrin-resistant Pediculus humanus capitis (Anoplura: Pediculidae) from Argentina. J Med Entomol. 2006; 43(5): 889–895. PubMed Abstract \| Publisher Full Text\n\nHeukelbach J, Canyon DV, Oliveira FA, et al.: In vitro efficacy of over-the-counter botanical pediculicides against the head louse Pediculus humanus var capitis based on a stringent standard for mortality assessment. Med Vet Entomol. 2008; 22(3): 264–272. PubMed Abstract \| Publisher Full Text\n\nToloza AC, Vassena C, Picollo MI: Ovicidal and adulticidal effects of monoterpenoids against permethrin-resistant human head lice, Pediculus humanus capitis. Med Vet Entomol. 2008; 22(4): 335–339. PubMed Abstract \| Publisher Full Text\n\nDi Campli E, Di Bartolomeo S, Delli Pizzi P, et al.: Activity of tea tree oil and nerolidol alone or in combination against Pediculus capitis (head lice) and its eggs. Parasitol Res. 2012; 111(5): 1985–1992. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbdel-Ghaffar F, Semmler M: Efficacy of neem seed extract shampoo on head lice of naturally infected humans in Egypt. Parasitol Res. 2007; 100(2): 329–332. PubMed Abstract \| Publisher Full Text\n\nIzri A, Uzzan B, Maigret M, et al.: Clinical efficacy and safety in head lice infection by Pediculus humanis capitis De Geer (Anoplura: Pediculidae) of a capillary spray containing a silicon-oil complex. Parasite. 2010; 17(4): 329–335. PubMed Abstract \| Publisher Full Text\n\nMehlhorn H, Abdel-Ghaffar F, Al-Rasheid KA, et al.: Ovicidal effects of a neem seed extract preparation on eggs of body and head lice. Parasitol Res. 2011; 109(5): 1299–1302. PubMed Abstract \| Publisher Full Text\n\nAbdel-Ghaffar F, Al-Quraishy S, Al-Rasheid KA, et al.: Efficacy of a single treatment of head lice with a neem seed extract: an in vivo and in vitro study on nits and motile stages. Parasitol Res. 2012; 110(1): 277–280. PubMed Abstract \| Publisher Full Text\n\nDwight Williams, Chana McIntyre and Chad Pilitere vs Warner-Lambert Consumer Healthcare, N.K.A. Pfizer, Inc., Bayer Corporation, Del Pharmaceutical, Inc., Del Laboratories, Inc. and Care Technologies, Inc. Cause No. B010023-C, Orange County, Texas. Information accessed 28th June 2013. Reference Source\n\nKatherine Mills, et al., Petitioners vs Warner Lambert Co., et al., Respondents. Cause No. 03-052, The Supreme Court of Texas. Information accessed 28th June 2013. Reference Source\n\nBurgess IF, Brunton ER, Burgess NA: Single application of 4% dimeticone liquid gel versus two applications of 1% permethrin creme rinse for treatment of head louse infestation: a randomised controlled trial. BMC Dermatology. 2013; 13: 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVander Stichele R, Dezeure EM, Bogaert MG: Systematic review of clinical efficacy of topical treatments for head lice. BMJ. 1995; 311(7005): 604–608. PubMed Abstract \| Free Full Text\n\nBurgess IF, Kay K, Brunton ER: Soya oil based shampoo superior to 0.5% permethrin lotion against head louse infestation: randomized clinical trial. Med Devices (Auckl). 2011; 4: 35–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurgess IF, Lee PN, Kay K, et al.: 1,2-octanediol, a novel surfactant, for treating head louse infestation: identification of activity, formulation, and randomised, controlled trials. PLoS One. 2012; 7(4): e35419. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurgess IF, Brown CM, Nair P: Individual demographic data collected from participants and lice found by detection combing post-treatment. F1000Research. 2014. Data Source" }	[ { "id": "5411", "date": "21 Jul 2014", "name": "Deon Canyon", "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 research was conducted well, and the article was well written. There is one grammatical error in the Outcomes section (paragraph 2) : \"In both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 first were [1st, 2nd...] stage nymphs that could only have originated from eggs not killed by insecticide.\" Also, nymphs can transfer so this conclusion is not accurate. \"However, success rates were significantly (p < 0.01) higher in people who had previously used a head louse treatment successfully, by an estimated factor of 1.88 (95% CI 1.13 to 3.11).\" So how many of the failures may have been due to inappropriately applying a treatment?The title and abstract are appropriate for the article, and provide a suitable summary. The experiment appears to have been conducted properly, with appropriate controls and data measurements, and the analysis is also adequate. The conclusions drawn from the study are both sensible and balanced, and the competing interests of the authors have been sufficiently disclosed.I'm just left wondering if any of the participant variables had an influence on the outcomes? If there was no effect, this should be stated.", "responses": [ { "c_id": "914", "date": "22 Jul 2014", "name": "Ian Burgess", "role": "Author Response", "response": "Thanks for finding the typographical error in Outcomes section (paragraph 2) : \"In both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 first were [1st, 2nd...] stage nymphs that could only have originated from eggs not killed by insecticide.\" that eluded everyone else reading it. It should have read, \"In both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 were first stage nymphs that could only have originated from eggs not killed by insecticide.\" While I do not disagree that young nymphs can transfer, this is a less common event than transfer of third stage nymphs and adults by some considerable factor of difference. Consequently, since relatively fewer adults were found in general it seems reasonable to assume that most, if not all, newly hatched first stage nymphs originated from eggs not killed by treatment. Consequently, I suggest a possible change to the text to read, \"In both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 were first stage nymphs the majority of which most likely originated from eggs not killed by insecticide.\", which would satisfy all possibilities.On the second point about how many treatment failures may have been due to inappropriate or incompetent application of treatment, the answer is we simply do not know. Firstly, common sense tells us that someone who has previously successfully negotiated the pitfalls of applying a head louse treatment is more likely to achieve a similar success on a subsequent occasion. Secondly, although the agency staff engaged by the sponsor were not experienced at performing head louse treatments so I do wonder how good they may have been (or not as the case may be) at detecting failures in application method. Thirdly, this is a risk you take when conducting pragmatic (real use by consumer) clinical studies. For most medications missing one dose is usually not the end of the world and does not usually affect the ultimate outcome. However, for a single dose application head louse study not applying the product properly is usually terminal from the perspective of efficacy, irrespective of how effective the product may or may not be, and in this case they were not as effective as the sponsors believed.So, in respect of participant variables, I do not doubt they influenced the outcome. The difficulty remains in determining which variable had what effect. In a wholly investigator run study, whether pragmatic or with investigator applied treatment, at least some of the variables are accountable and can be analysed accordingly. However, in this case I don't think anyone could have done much to greatly affect the overall outcome." } ] }, { "id": "5970", "date": "28 Aug 2014", "name": "Maria Inés Picollo", "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 clinical study to assess the comparative effectiveness of three head louse treatments when are used by the public. For this purpose, the treatments selected were 0.5% d-phenothrin mousse, 0.2% d-phenothrin lotion, and wet-combing with conditioner. The commercial products were applied according to manufacturer's instructions (30 minutes for 0,5% phenothrin mousse and 2 hours for 0,2% phenothrin lotion), and wet-combing according to the Bug-Busting Community Hygiene Concern, London (fine toothed plastic comb with conditioner, repeated at 3–4 day intervals for 2 weeks).I consider that the manuscript is within the scope of the journal, the abstract is according to the objectives and results of the manuscript: The methodology, results and discussion are carefully written and adequate to the objectives of the study.One of the relevant aspects of the work is that reinforces the need for the effectiveness of commercial products must be warranted by the health authorities, due to the importance of pediculosis in school-aged children both in developed and developing countries. An unexpected and surprising result is the low effectiveness of the three methods evaluated. Although pyrethroid resistance and inadequate formulation of the active ingredients surely contribute to the low effectiveness of the phenothrin-based formulations (20% and 21,5% success for mousse and lotion respectively), the low effectiveness of wet-combing (19,15 success) has not been previously reported. This result is worrying considering that the use of a fine comb for removing lice and nits was proposed as a relevant tool both in the diagnosis of infestations and as part of an integrated control strategy of head lice. I agree with the authors in their first conclusion about the limitations of laboratory tests: ¨bioassay tests conducted in a laboratory, whether using laboratory reared insects or even wild collected ones, can only be indicative of efficacy for a formulation and it is unknown for a treatment to perform poorly in vitro yet be effective in vivo. ….. even ex vivo screens, using insects recently collected from the wild, may be only partially representative and several replicate tests should be performed using insects from geographically separated locations to ensure that the outcome is not obtained either by chance or due to some happenstance of physiological difference in the insects from that location¨. But as the authors surely know and probably should clarify in the paper, the main purpose of laboratory testing is the ¨comparative-evaluation¨ of the effectiveness of pure compounds or formulations. Thus, the parallel and simultaneous evaluation of different compounds and/or formulations made in standardized conditions, is a reliable result of the comparative activity of the products evaluated.Obviously laboratory tests are faster and cheaper than clinical trials, and represent an excellent tool for a first selection of the compound or product to be developed (a pure compound without pediculicide activity in laboratory tests, will not be effective in a clinical trial).", "responses": [] } ]	1	https://f1000research.com/articles/3-158
https://f1000research.com/articles/3-155/v1	08 Jul 14	{ "type": "Review", "title": "Vitamin D as a regulator of steroidogenic enzymes", "authors": [ "Johan Lundqvist" ], "abstract": "During the last decades, the outlook on vitamin D has widened, from being a vitamin solely involved in bone metabolism and calcium homeostasis, to being a multifunctional hormone known to affect a broad range of physiological processes. The aim of this review is to summarize the research on vitamin D as a regulator of steroidogenic enzymes. Steroid hormones exert a wide range of physiological responses, including functions in the immune system, protein and carbohydrate metabolism, water and salt balance, reproductive system and development of sexual characteristics. The balance of sex hormones is also of importance in the context of breast and prostate cancer. Steroid hormones are synthesized in steroidogenic tissues such as the adrenal cortex, breast, ovaries, prostate and testis, either from cholesterol or from steroidogenic precursors secreted from other steroidogenic tissues. The hormonally active form of vitamin D has been reported to act as a regulator of a number of enzymes involved in the regulation of steroid hormon production, and thereby the production of both adrenal steroid hormones and sex hormones. The research reviewed in the article has in large part been performed in cell culture based experiments and laboratory animal experiments, and the physiological role of the vitamin D mediated regulation of steroidogenic enzyme need to be further investigated.", "keywords": [ "Vitamin D was discovered in the early 20th century when it was described that rickets could be cured by sunlight or cod liver oil. This fat-soluble vitamin was the fourth vitamin described and therefore designated as vitamin D. Vitamin D was found to regulate intestinal absorption and renal reabsorption of calcium and bone metabolism. Later research has demonstrated that vitamin D can be obtained from the diet or be de novo synthesized from 7-dehydrocholesterol. Further", "the active form of vitamin D", "1α", "25-dihydroxyvitamin D3", "interacts with a receptor in the target cells", "showing that vitamin D should be biochemically characterized as a hormone rather than as a vitamin1." ], "content": "Vitamin D - a multifunctional hormone\n\nVitamin D was discovered in the early 20th century when it was described that rickets could be cured by sunlight or cod liver oil. This fat-soluble vitamin was the fourth vitamin described and therefore designated as vitamin D. Vitamin D was found to regulate intestinal absorption and renal reabsorption of calcium and bone metabolism. Later research has demonstrated that vitamin D can be obtained from the diet or be de novo synthesized from 7-dehydrocholesterol. Further, the active form of vitamin D, 1α,25-dihydroxyvitamin D3, interacts with a receptor in the target cells, showing that vitamin D should be biochemically characterized as a hormone rather than as a vitamin1.\n\nDuring the last decades, the outlook on vitamin D has widened, from being a vitamin solely involved in bone metabolism and calcium homeostasis, to being a multifunctional hormone known to affect a broad range of physiological processes. This includes effects on the immune system, brain and fetal development, insulin secretion, cancer, apoptosis, cell proliferation and differentiation as well as the cardiovascular system via the vitamin D receptor (VDR)1–4. The vitamin D receptor is widely expressed and it has been suggested that 1α,25-dihydroxyvitamin D3 may have other roles yet undiscovered3,5,6. The aim of this paper is to review the literature on effects of vitamin D and vitamin D analogs on steroidogenic enzymes.\n\nThere are two forms of vitamin D, vitamin D2 (ergocalciferol) and vitamin D3 (cholecalciferol). Ergocalciferol is synthesized in plants, yeast and fungi while cholecalciferol is synthesized in animals. Vitamin D3 is synthesized in the skin from 7-dehydrocholesterol upon exposure to UV-B radiation. Vitamin D3 is then bioactivated in two subsequent steps to gain the biologically active form of vitamin D (Figure 1). In the first step, vitamin D3 is 25-hydroxylated to 25-hydroxyvitamin D3 (calcidiol). 25-Hydroxylation of vitamin D is a reaction that can be catalyzed by the mitochondrial CYP27A1 and the microsomal CYP3A4, CYP2R1 and CYP2J2 in humans. 25-Hydroxylation of vitamin D is mainly performed in the liver and calcidiol is then excreted into the circulation. Calcidiol is converted to 1α,25-dihydroxyvitamin D3 (calcitriol) by 1α-hydroxylation, mainly performed in the kidneys. The principal human 1α-hydroxylase for 25-hydroxyvitamin D3 is CYP27B1. The 1α,25-dihydroxyvitamin D3 produced is excreted into the circulation and acts as a hormone1,3,4,7–9.\n\nExtrarenal 1α-hydroxylation of 25-hydroxyvitamin D3 has been reported for a wide range of tissues, including colon, brain, mammary tissue, breast, pancreatic islets, parathyroid glands, placenta, prostate and keratinocytes3,7. These findings suggest that 1α,25-dihydroxyvitamin D3 may be produced locally and act in an intracrine or paracrine fashion. It may be speculated that the local concentration of 1α,25-dihydroxyvitamin D3 in these tissues could be higher than the circulating levels.\n\nThe normal serum level of calcidiol is 50–100 nM and for calcitriol 50–125 pM3. Calcitriol is the most potent form of vitamin D even though calcidiol can exert some biological effects as well. The circulating levels of 1α,25-dihydroxyvitamin D3 is tightly regulated via a feed-back mechanism where 1α,25-dihydroxyvitamin D3 downregulates the expression of CYP27B1 and upregulates the expression of CYP24A18.\n\nBoth calcidiol and calcitriol is metabolized by CYP24A1 to the less active compounds 24,25-dihydroxyvitamin D3 and 1α,24,25-trihydroxyvitamin D3 respectively10. It has recently been reported that CYP11A1 can catalyze the production of 20-hydroxyvitamin D3 from vitamin D3 and the production of 1α,20-dihydroxyvitamin D3 from 1α-hydroxyvitamin D311–15. Both these metabolites have been reported to exert biological effects on cell differentiation and gene expression in a way resembling the one of 1α,25-dihydroxyvitamin D311,12. The physiological role, if any, of this CY11A1-mediated metabolism of vitamin D remains to be clarified.\n\nThe bioactivated form of vitamin D alters the gene expression of a large number of genes. It is well known that 1α,25-dihydroxyvitamin D3 can act either to increase the gene expression or decrease the gene expression, depending on the gene in question. For example, 1α,25-dihydroxyvitamin D3 increases the gene expression of CYP24A1 while it increases the gene expression of CYP27B1. Both these effects of the hormonally active form of vitamin D are mediated via a VDR dependent mechanism. The mechanism for 1α,25-dihydroxyvitamin D3-mediated induction of gene expression is well known and based on the interaction between the 1α,25-dihydroxyvitamin D3-activated VDR and a vitamin D responsive element (VDRE) in the gene promoter. These positive VDRE (pVDRE) consist of a hexameric direct repeat of the consensus sequence 5´-RGKTCA (R=A or G, K=G or T)16. The two half sites are separated by a three nucleotide spacer. The ligand-activated VDR-RXR complex interacts with the pVDRE and acts as a transcription factor to increase the transcriptional rate by recruiting coactivators3.\n\nHowever, the mechanism for 1α,25-dihydroxyvitamin D3-mediated downregulation of gene expression has in large part remained unclear17. Studies have only been performed for a few genes, regarding the molecular mechanism for the vitamin D-mediated downregulation of gene expression. For these genes, it has been suggested that the mechanism could include recruitment or displacement of corepressors, such as VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF), from the promoter sequence. Further, it has been proposed that epigenetic changes such as histone deacetylation and DNA methylation might be involved in the mechanism18–23. The negative vitamin D response elements (nVDRE) described show a very low level of similarity to the pVDRE described21. Furthermore, it has been proposed that the mechanism for vitamin D-mediated downregulation of gene expression is not based on a direct interaction between the ligand-activated VDR and the promoter sequence, but rather an in-direct interaction via comodulator VDIR18.\n\nRecently, it has been shown in genome-wide ChIP-seq experiments that VDR has a large number of binding sites throughout the genome24–26. These binding sites have been found to be located predominantly within introns and intergenic regions and often far away from the transcriptional start site26. The physiological role of these binding sites remains to be elucidated.\n\nVitamin D has also been reported to exert rapid effects only seconds or minutes after treatment. Due to the quick response, it has been suggested that these effects are non-genomic and mediated by membrane-bound receptors27.\n\n\nSteroid hormone synthesis\n\nAll steroid hormones are synthesized from the common precursor cholesterol, which can be obtained from the diet or de novo synthesized from acetyl CoA. The production of steroid hormones is regulated via a number of enzymes of which a majority belongs to the cytochrome P450 (CYP) superfamily.\n\nSteroid hormones exert a wide range of physiological responses, including functions in the immune system, protein and carbohydrate metabolism, water and salt balance, reproductive system and development of sexual characteristics. Steroid hormones are synthesized in steroidogenic tissues such as the adrenal cortex, breast, ovaries, prostate and testis, either from cholesterol or from steroidogenic precursors secreted from other steroidogenic tissues.\n\nThe adrenal cortex produces steroid hormones such as aldosterone, corticosterone, cortisol, dehydroepiandrosterone (DHEA) and androstenedione. An overview of steroids and enzyme-catalyzed reactions in the adrenal steroidogenesis is shown in Figure 2. The adrenal steroidogenesis is quantitatively regulated by the transcription of CYP11A1 (cholesterol side-chain cleavage enzyme) and the activity of steroidogenic acute regulatory protein (StAR).\n\nThere are three adrenocortical zones, each with a distinct role in the production of steroid hormones; zona glomerulosa produces mineralocorticoids (e.g. aldosterone), zona fasciculata produces glucocorticoids (e.g. cortisol) and zona reticularis is the point of synthesis for adrenal androgens (e.g. DHEA). The qualitative regulation of adrenal steroidogenesis, determining which type of steroid that will be produced, is performed by the transcription and activity of CYP17A1. CYP17A1 catalyzes two different reactions, namely the 17α-hydroxylation and the 17,20-lyase reaction. The expression and activity of CYP17A1 differs between the three adrenal zones28–30. CYP17A1 is not expressed in zona glomerulosa leading to the production of mineralocorticoids. In zona fasciculata, CYP17A1 is expressed and catalyzing 17α-hydroxylation, but not the 17,20-lyase activity, leading to glucocorticoid production. In zona reticularis, CYP17A1 catalyzes both 17α-hydroxylation and 17,20-lyase activities and adrenal androgens are therefore the main product of adrenal steroidogenesis in this adrenal zone28. To control the production of steroid hormones, it is essential to regulate the two activities of CYP17A1 separately. The 17α-hydroxylase activity of CYP17A1 is regulated via the gene expression of CYP17A1. On the other hand, the 17,20-lyase activity of CYP17A1 is regulated via posttranscriptional mechanisms28–34. The adrenal zone-specific activities of CYP17A1 are summarized in Figure 3.\n\nThe 17,20-lyase activity has been reported to be regulated via three posttranscriptional mechanism; the abundance of P450 oxidoreductace (POR)35,36, allosteric action of cytochrome b537 and serine phosphorylation of CYP17A138–41. In 1972, Zachman et al.42 described the first case of isolated 17,20-lyase deficiency, which is a rare condition. 17,20-Lyase deficiency could be a result of a mutated cytochrome b5, according to a recent report43. CYP17A1 deficiency, or impaired CYP17A1 activity due to altered posttranscriptional mechanisms, may lead to hypertension, hypokalemia and impaired development of sexual characteristics due to decreased production of adrenal androgens. Patients that are genetically male often present complete male pseudohermaphroditism while female patients may be infertile44–46.\n\nIt has been reported that CYP17A1 strongly prefers 17α-hydroxypregnenolone over 17α-hydroxyprogesterone as a substrate for 17,20-lyase activity in humans28,47. 17α-Hydroxyprogesterone may, however, be a substrate for CYP17A1 in other species47.\n\nCYP21A2 is a steroid 21-hydroxylase catalyzing the production of deoxycorticosterone and 11-deoxycortisol, which are precursors for the production of corticosterone, aldosterone and cortisol. 21-Hydroxylase deficiency may lead to severe conditions such as congenital adrenal hyperplasia and Addison´s disease48. CYP21A2 is exclusively expressed in the adrenal cortex28. Therefore, glucocorticoids and mineralocorticoids cannot be synthesized in other tissues than the adrenal cortex.\n\nSex hormones are mainly produced in the gonads, breast and prostate. The sex hormones in these tissues are produced either by in situ synthesis from cholesterol or by enzyme catalyzed conversion of DHEA or androstenedione excreted to the circulation from the adrenal cortex.\n\nThe sex hormones are divided into two groups; androgens and estrogens. Androgens such as testosterone and 5α-dihydrotestosterone (DHT) are produced via reactions catalyzed by 17β-hydroxysteroid dehydrogenase (17β-HSD) and 5α-reductase. Estrogens, such as 17β-estradiol (estradiol), are produced by aromatization of androgenic precursors, a reaction catalyzed by CYP19A1 (aromatase). Hence, the tissue-selective expression and activity of 5α-reductase and aromatase regulates the production of androgens and estrogens49,51. For some of these steroids, it remains unclear if they act as estrogens or as androgens or if they are inactive metabolites52. To fully understand the estrogenic and/or androgenic signaling exerted by these steroids is of utmost importance in research on eg. breast and prostate carcinogenesis.\n\nIt is crucial to regulate the levels of sex hormones in order to achieve normal gonadal development. Abnormally high levels of estrogens and androgens are associated with increased risk for breast cancer and prostate cancer, respectively50,53–55.\n\n\nVitamin D and analogs as regulators of steroidogenic enzymes\n\nA link between vitamin D and the adrenal steroidogenesis has been proposed in a few very early reports56,57. In a paper from 1959, De Toni et al.56 describes several clinical cases involving children with rickets having changes in urinary 17-ketosteroid levels. It is suggested in the paper that the altered steroid production may involve some action of vitamin D. Furthermore, the authors propose that disturbances in steroid metabolism might be due to sensitivity or resistance of organisms to antirachitic vitamins57. More recently, a case has been described where a woman with osteomalacia was reported to have elevated serum levels of aldosterone and simultaneously low levels of 25-hydroxyvitamin D58. After 24 months of treatment with vitamin D, the condition was normalized.\n\nRecently, Lundqvist et al.59 investigated the effects of 1α,25-dihydroxyvitamin D3 on the adrenal steroidogenesis. We studied the effects of 1α,25-dihydroxyvitamin D3 on the gene expression of key steroidogenic enzymes, the enzyme activity and the hormone production. The study was performed in the human adrenocortical carcinoma cell line NCI-H295R. We found that the mRNA levels of three key enzymes in the adrenal steroidogenesis, CYP11A1, CYP17A1 and CYP21A2 were altered by 1α,25-dihydroxyvitamin D3 treatment. CYP11A1 and CYP17A1 mRNA levels were upregulated after vitamin D treatment, while CYP21A2 mRNA level was suppressed by the same treatment. No significant changes were observed in the mRNA levels of CYP11B1, CYP11B2 and 3βHSD. Further, we found that 1α,25-dihydroxyvitamin D3 treatment decreased the production of corticosterone, androstenedione, dehydroepiandrosterone (DHEA) and DHEA-sulfate (DHEA-S), while the production of aldosterone and cortisol was unaltered. Moreover, we measured the enzyme activity of CYP21A2 and CYP17A1 by adding a known amount of substrate to the cell culture and measuring the turnover of substrate to product. In resemblance with the effects on mRNA level, CYP21A2 enzyme activity was suppressed by 1α,25-dihydroxyvitamin D3. CYP17A1 catalyzes two separate reactions, the 17α-hydroxylation and the 17,20-lyase reaction. We found that treatment with vitamin D resulted in increased 17α-hydroxylation activity of CYP17A1, but in decreased 17,20-lyase activity of CYP17A1.\n\nThe regulation of the two activities of CYP17A1 uses two principally different mechanisms. The 17α-hydroxylase activity is regulated by alterations of the gene expression of CYP17A1. The 17,20-lyase activity, on the other hand, is regulated via posttranscriptional mechanisms28,31–34. It has been reported that the 17,20-lyase activity is regulated via three mechanisms; the abundance of P450 oxidoreductase (POR)35,36, allosteric action of cytochrome b537 and serine phosphorylation of CYP17A138–41. The discrepancy between the 1α,25-dihydroxyvitamin D3-mediated increase in expression of CYP17A1 mRNA and the suppression of 17,20-lyase activity could be a result of posttranscriptional mechanisms affected by 1α,25-dihydroxyvitamin D3.\n\nIn a subsequent report22, we investigated the molecular mechanism for the effect of 1α,25-dihydroxyvitamin D3 on CYP21A2 gene expression. We found that 1α,25-Dihydroxyvitamin D3 altered the promoter activity of CYP21A2 via a mechanism involving VDR and a vitamin D response element in the CYP21A2 promoter. Further, we found that the mechanism included interaction of the comodulators VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF) to the gene promoter.\n\nChatterjee and collaborators60,61 have reported that ligand-activated VDR upregulates the expression of sulfotransferase 2A1 (SULT2A1), an enzyme that catalyzes the conversion of dehydroepiandrosterone (DHEA) to dehydroepiandrosterone-sulfate (DHEA-S).\n\nThe production of sex hormones is regulated by multiple enzymes. Vitamin D has been reported to affect the expression and activity of several of these enzymes.\n\n17β-hydroxysteroiddehydrogenases. Wang and Tuohimaa62 have reported that 1α,25-dihydroxyvitamin D3 upregulates the mRNA level for 17β-hydroxysteroid dehydrogenases (17β-HSD) type 2, 4 and 5 in cell lines derived from human prostate. In keratinocytes, Hughes et al.63 have reported that 1α,25-dihydroxyvitamin D3 stimulates the expression of 17β-HSD type 1 and 2.\n\nAromatase. It has been shown that 1α,25-dihydroxyvitamin D3 alters the aromatase activity in placental cells64,65, prostate cells66 and osteoblasts67,68. Kinuta et al.69 have reported that vitamin D receptor null mutant mice have a decreased aromatase activity in the ovary, testis and epididymis. Recently, it was reported that 1α,25-dihydroxyvitamin D3 alters the gene expression of aromatase in a tissue-selective manner70. In breast cancer cell lines, vitamin D treatment resulted in decreased aromatase gene expression, while the same treatment increased the aromatase gene expression in osteosarcoma cell lines. 1α,25-Dihydroxyvitamin D3 has therefore been proposed to be a tissue-selective aromatase modulator70.\n\nWe have investigated the effects of 1α,25-dihydroxyvitamin D3 on the aromatase gene expression and estradiol production in human breast carcinoma MCF-7 cells, human adrenocortical carcinoma NCI-H295R cells and human prostate cancer LNCaP cells71. We found that the hormonally active form of vitamin D altered the estrogen and androgen metabolism in a cell line specific manner. Aromatase gene expression and estradiol production was found to be decreased in breast cancer cells, while the androgen production was markedly increased in the same cell line. 1α,25-dihydroxyvitamin D3 was found to increase the aromatase gene expression and decrease dihydrotestosterone production in adrenocortical cells. In prostate cancer cells, aromatase gene expression was found to be increased after 1α,25-dihydroxyvitamin D3 treatment. Furthermore, we studied the effects of 1α,25-dihydroxyvitamin D3 on three different aromatase promoters, and found that the transcriptional rate of these promoters were affected by 1α,25-dihydroxyvitamin D3 in a cell line-specific manner.\n\nIn a subsequent study, we have shown that the substance EB1089 (a vitamin D analog with decreased hypercalcemic effect) is able to inhibit the aromatase gene expression by dissociation of comodulator WSTF from the CYP19A1 promoter in human breast cancer MCF-7 cells23. Furthermore, 1α,25-dihydroxyvitamin D3 and analogs have been reported to alter sex hormone signaling by suppressing the expression of estrogen receptor α72–77. It has also been reported that vitamin D deficiency alters reproductive functions in both male and female rats, indicating that vitamin D may affect sex hormone signaling78,79.\n\n\nConcluding remarks\n\nIn conclusion, 1α,25-dihydroxyvitamin D3 has been shown to regulate the gene expression and enzyme activity of a number of steroidogenic enzymes, and the corresponding hormone production. The physiological role of these vitamin D effects remains to be elucidated in many cases, especially for the adrenal steroidogenic enzymes. More research has been conducted on the physiological role and potential medical use of the effects of vitamin D on estrogen production and action in breast cancer. Cell culture based experiments and laboratory animal experiments has clearly shown that 1α,25-dihydroxyvitamin D3 is able to decrase the production and action of estrogens in breast cancer models, and also the growth of breast cancer xenotumors70,72–75,80. Based on these findings, vitamin D or analogs has been proposed to be of potential use as an anti-cancer agent76,81–84 and a large number of clinical trials (e.g. ClinicalTrials.gov identifiers NCT00656019, NCT01472445, NCT01965522, NCT01948128, NCT01787409, and NCT01097278) are currently being conducted using vitamin D or analogs in different settings to investigate the potential use in the prevention or treatment of breast cancer.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author is grateful for financial support from the Swedish Academy of Pharmaceutical Sciences and the Swedish Society for Medical 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\nPlum LA, DeLuca HF: Vitamin D, disease and therapeutic opportunities. Nat Rev Drug Discov. 2010; 9(12): 941–55. PubMed Abstract \| Publisher Full Text\n\nDusso AS, Brown AJ, Slatopolsky E: Vitamin D. Am J Physiol Renal Physiol. 2005; 289(1): F8–28. PubMed Abstract \| Publisher Full Text\n\nNorman AW: From vitamin D to hormone D: fundamentals of the vitamin D endocrine system essential for good health. Am J Clin Nutr. 2008; 88(2): 491S–499S. PubMed Abstract\n\nHolick MF: Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and cardiovascular disease. Am J Clin Nutr. 2004; 80(6 Suppl): 1678S–88S. PubMed Abstract\n\nJones G, Strugnell SA, DeLuca HF: Current understanding of the molecular actions of vitamin D. Physiol Rev. 1998; 78(4): 1193–1231. PubMed Abstract\n\nDeLuca HF: Overview of general physiologic features and functions of vitamin D. Am J Clin Nutr. 2004; 80(6 Suppl): 1689S–96S. PubMed Abstract\n\nHolick MF: Vitamin D: A millenium perspective. J Cell Biochem. 2003; 88(2): 296–307. PubMed Abstract \| Publisher Full Text\n\nProsser DE, Jones G: Enzymes involved in the activation and inactivation of vitamin D. Trends Biochem Sci. 2004; 29(12): 664–673. PubMed Abstract \| Publisher Full Text\n\nKarlgren M, Miura S, Ingelman-Sundberg M: Novel extrahepatic cytochrome P450s. Toxicol Appl Pharmacol. 2005; 207(2 Suppl): 57–61. PubMed Abstract \| Publisher Full Text\n\nDeeb KK, Trump DL, Johnson CS: Vitamin D signalling pathways in cancer: potential for anticancer therapeutics. Nat Rev Cancer. 2007; 7(9): 684–700. PubMed Abstract \| Publisher Full Text\n\nZbytek B, Janjetovic Z, Tuckey RC, et al.: 20-Hydroxyvitamin D3, a product of vitamin D3 hydroxylation by cytochrome P450scc, stimulates keratinocyte differentiation. J Invest Dermatol. 2008; 128(9): 2271–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTuckey RC, Janjetovic Z, Li W, et al.: Metabolism of 1alpha-hydroxyvitamin D3 by cytochrome P450scc to biologically active 1alpha,20-dihydroxyvitamin D3. J Steroid Biochem Mol Biol. 2008; 112(4–5): 213–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlominski A, Semak I, Wortsman J, et al.: An alternative pathway of vitamin D2 metabolism. Cytochrome P450scc (CYP11A1)-mediated conversion to 20-hydroxyvitamin D2 and 17,20-dihydroxyvitamin D2. FEBS J. 2006; 273(13): 2891–2901. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlominski A, Semak I, Zjawiony J, et al.: Enzymatic metabolism of ergosterol by cytochrome P450scc to biologically active 17alpha,24-dihydroxyergosterol. Chem Biol. 2005; 12(8): 931–939. PubMed Abstract \| Publisher Full Text\n\nSlominski A, Semak I, Zjawiony J, et al.: The cytochrome P450scc system opens an alternate pathway of vitamin D3 metabolism. FEBS J. 2005; 272(16): 4080–4090. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMatilainen M, Malinen M, Saavalainen K, et al.: Regulation of multiple insulin-like growth factor binding protein genes by 1alpha,25-dihydroxyvitamin D3. Nucleic Acids Res. 2005; 33(17): 5521–5532. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPike JW, Meyer MB: The vitamin D receptor: new paradigms for the regulation of gene expression by 1,25-dihydroxyvitamin D(3). Endocrinol Metab Clin North Am. 2010; 39(2): 255–269. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFujiki R, Kim MS, Sasaki Y, et al.: Ligand-induced transrepression by VDR through association of WSTF with acetylated histones. EMBO J. 2005; 24(22): 3881–3894. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMurayama A, Kim MS, Yanagisawa J, et al.: Transrepression by a liganded nuclear receptor via a bHLH activator through co-regulator switching. EMBO J. 2004; 23(7): 1598–1608. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim MS, Fujiki R, Kitagawa H, et al.: 1alpha,25(OH)2D3-induced DNA methylation suppresses the human CYP27B1 gene. Mol Cell Endocrinol. 2007; 265–266: 168–173. PubMed Abstract \| Publisher Full Text\n\nKim MS, Fujiki R, Murayama A, et al.: 1alpha,25(OH)2D3-induced transrepression by vitamin D receptor through E-box-type elements in the human parathyroid hormone gene promoter. Mol Endocrinol. 2007; 21(2): 334–342. PubMed Abstract \| Publisher Full Text\n\nLundqvist J, Wikvall K, Norlin M: Vitamin D-mediated regulation of CYP21A2 transcription - a novel mechanism for vitamin D action. Biochim Biophys Acta - General Subjects. 2012; 1820(10): 1553–1559. PubMed Abstract \| Publisher Full Text\n\nLundqvist J, Hansen SK, Lykkesfeldt AE, et al.: Vitamin D analog EB1089 inhibits aromatase expression by dissociation of comodulator WSTF from the CYP19A1 promoter-a new regulatory pathway for aromatase. Biochim Biophys Acta. 2013; 1833(1): 40–47. PubMed Abstract \| Publisher Full Text\n\nTuoresmäki P, Väisänen S, Neme A, et al.: Patterns of Genome-Wide VDR Locations. PLoS One. 2014; 9(4): e96105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeyer MB, Pike JW: Corepressors (NCoR and SMRT) as well as coactivators are recruited to positively regulated 1α,25-dihydroxyvitamin D3-responsive genes. J Steroid Biochem Mol Biol. 2013; 136: 120–124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPike JW, Meyer MB: Fundamentals of vitamin D hormone-regulated gene expression. J Steroid Biochem Mol Biol. In press. 2013. PubMed Abstract \| Publisher Full Text\n\nNorman AW, Okamura WH, Bishop JE, et al.: Update on biological actions of 1alpha,25(OH)2-vitamin D3 (rapid effects) and 24R,25(OH)2-vitamin D3. Mol Cell Endocrinol. 2002; 197(1–2): 1–13. PubMed Abstract \| Publisher Full Text\n\nMiller WL: Steroidogenic Enzymes. Endocr Dev. 2008; 13: 1–18. PubMed Abstract \| Publisher Full Text\n\nMiller WL: Androgen synthesis in adrenarche. Rev Endocr Metab Disord. 2009; 10(1): 3–17. PubMed Abstract \| Publisher Full Text\n\nRainey WE, Nakamura Y: Regulation of the adrenal androgen biosynthesis. J Steroid Biochem Mol Biol. 2008; 108(3–5): 281–286. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNguyen AD, Conley AJ: Adrenal androgens in humans and nonhuman primates: production, zonation and regulation. Endocr Dev. 2008; 13: 33–54. PubMed Abstract \| Publisher Full Text\n\nMiller WL, Huang N, Agrawal V, et al.: Genetic variation in human P450 oxidoreductase. Mol Cell Endocrinol. 2009; 300(1–2): 180–184. PubMed Abstract \| Publisher Full Text\n\nMiller WL, Auchus RJ, Geller DH: The regulation of 17,20 lyase activity. Steroids. 1997; 62(1): 133–142. PubMed Abstract \| Publisher Full Text\n\nHall PF: Cytochrome P-450 C21scc: one enzyme with two actions: hydroxylase and lyase. J Steroid Biochem Mol Biol. 1991; 40(4–6): 527–532. PubMed Abstract \| Publisher Full Text\n\nLin D, Black SM, Nagahama Y, et al.: Steroid 17 alpha-hydroxylase and 17,20-lyase activities of P450c17: contributions of serine106 and P450 reductase. Endocrinology. 1993; 132(6): 2498–2506. PubMed Abstract \| Publisher Full Text\n\nYanagibashi K, Hall PF: Role of electron transport in the regulation of the lyase activity of C21 side-chain cleavage P-450 from porcine adrenal and testicular microsomes. J Biol Chem. 1986; 261(18): 8429–8433. PubMed Abstract\n\nAuchus RJ, Lee TC, Miller WL: Cytochrome b5 augments the 17,20-lyase activity of human P450c17 without direct electron transfer. J Biol Chem. 1998; 273(6): 3158–3165. PubMed Abstract \| Publisher Full Text\n\nPandey AV, Mellon SH, Miller WL: Protein phosphatase 2A and phosphoprotein SET regulate androgen production by P450c17. J Biol Chem. 2003; 278(5): 2837–2844. PubMed Abstract \| Publisher Full Text\n\nPandey AV, Miller WL: Regulation of 17,20 lyase activity by cytochrome b5 and by serine phosphorylation of P450c17. J Biol Chem. 2005; 280(14): 13265–13271. PubMed Abstract \| Publisher Full Text\n\nZhang LH, Rodriguez H, Ohno S, et al.: Serine phosphorylation of human P450c17 increases 17,20-lyase activity: implications for adrenarche and the polycystic ovary syndrome. Proc Natl Acad Sci U S A. 1995; 92(23): 10619–10623. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang YH, Tee MK, Miller WL: Human Cytochrome P450c17: single step purification and phosphorylation of serine 258 by protein kinase a. Endocrinology. 2010; 151(4): 1677–1684. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZachmann M, Vollmin JA, Hamilton W, et al.: Steroid 17,20-desmolase deficiency: a new cause fo male pseudohermaphroditism. Clin Endocrinol (Oxf). 1972; 1(4): 369–385. PubMed Abstract \| Publisher Full Text\n\nKok RC, Timmerman MA, Wolffenbuttel KP, et al.: Isolated 17,20-lyase deficiency due to the cytochrome b5 mutation W27X. J Clin Endocrinol Metab. 2010; 95(3): 994–999. PubMed Abstract \| Publisher Full Text\n\nNew MI: Inborn errors of adrenal steroidogenesis. Mol Cell Endocrinol. 2003; 211(1–2): 75–83. PubMed Abstract \| Publisher Full Text\n\nMiller WL: Disorders of androgen synthesis--from cholesterol to dehydroepiandrosterone. Med Princ Pract. 2005; 14(Suppl 1): 58–68. PubMed Abstract \| Publisher Full Text\n\nFlück CE, Pandey AV: Clinical and biochemical consequences of p450 oxidoreductase deficiency. Endocr Dev. 2011; 20: 63–79. PubMed Abstract\n\nGilep AA, Sushko TA, Usanov SA: At the crossroads of steroid hormone biosynthesis: the role, substrate specificity and evolutionary development of CYP17. Biochim Biophys Acta (BBA) - Proteins & Proteomics. 2011; 1814(1): 200–209. PubMed Abstract \| Publisher Full Text\n\nKrone N, Arlt W: Genetics of congenital adrenal hyperplasia. Best Pract Res Clin Endocrinol Metab. 2009; 23(2): 181–192. PubMed Abstract \| Publisher Full Text\n\nEllem SJ, Risbridger GP: Aromatase and regulating the estrogen:androgen ratio in the prostate gland. J Steroid Biochem Mol Biol. 2010; 118(4–5): 246–251. PubMed Abstract \| Publisher Full Text\n\nRisbridger GP, Davis ID, Birrell SN, et al.: Breast and prostate cancer: more similar than different. Nat Rev Cancer. 2010; 10(3): 205–212. PubMed Abstract \| Publisher Full Text\n\nSimpson ER, Clyne C, Rubin G, et al.: Aromatase--a brief overview. Annu Rev Physiol. 2002; 64: 93–127. PubMed Abstract \| Publisher Full Text\n\nLardy H, Marwah A, Marwah P: C(19)-5-ene steroids in nature. Vitam Horm. 2005; 71: 263–299. PubMed Abstract \| Publisher Full Text\n\nAli S, Coombes RC: Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer. 2002; 2(2): 101–112. PubMed Abstract \| Publisher Full Text\n\nScher HI, Sawyers CL: Biology of progressive, castration-resistant prostate cancer: directed therapies targeting the androgen-receptor signaling axis. J Clin Oncol. 2005; 23(32): 8253–8261. PubMed Abstract \| Publisher Full Text\n\nBalk SP, Knudsen KE: AR, the cell cycle, and prostate cancer. Nucl Recept Signal. 2008; 6: e001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Toni E Jr, Nordio S: The relationship between calcium-phosphorus metabolism, the 'Krebs cycle' and steroid metabolism. Arch Dis Child. 1959; 34: 371–382. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNatanson AO, Chuvaev AV: [The effect of a rachitogenic regimen on the function of the rat adrenal cortex]. Vopr Pitan. 1970; 29(3): 46–47. PubMed Abstract\n\nTaylor HC, Elbadawy EH: Renal tubular acidosis type 2 with Fanconi's syndrome, osteomalacia, osteoporosis, and secondary hyperaldosteronism in an adult consequent to vitamin D and calcium deficiency: effect of vitamin D and calcium citrate therapy. Endocr Pract. 2006; 12(5): 559–567. PubMed Abstract \| Publisher Full Text\n\nLundqvist J, Norlin M, Wikvall K: 1alpha,25-Dihydroxyvitamin D3 affects hormone production and expression of steroidogenic enzymes in human adrenocortical NCI-H295R cells. Biochim Biophys Acta (BBA) - Molecular and Cell Biology of Lipids. 2010; 1801(9): 1056–1062. PubMed Abstract \| Publisher Full Text\n\nChatterjee B, Echchgadda I, Song C: Vitamin D receptor regulation of the steroid/bile acid sulfotransferase SULT2A1. Methods Enzymol. 2005; 400: 165–191. PubMed Abstract \| Publisher Full Text\n\nEchchgadda I, Song CS, Roy AK, et al.: Dehydroepiandrosterone sulfotransferase is a target for transcriptional induction by the vitamin D receptor. Mol Pharmacol. 2004; 65(3): 720–729. PubMed Abstract \| Publisher Full Text\n\nWang JH, Tuohimaa P: Regulation of 17beta-hydroxysteroid dehydrogenase type 2, type 4 and type 5 by calcitriol, LXR agonist and 5alpha-dihydrotestosterone in human prostate cancer cells. J Steroid Biochem Mol Biol. 2007; 107(1–2): 100–105. PubMed Abstract \| Publisher Full Text\n\nHughes SV, Robinson E, Bland R, et al.: 1,25-dihydroxyvitamin D3 regulates estrogen metabolism in cultured keratinocytes. Endocrinology. 1997; 138(9): 3711–3718. PubMed Abstract \| Publisher Full Text\n\nBarrera D, Avila E, Hernández G, et al.: Estradiol and progesterone synthesis in human placenta is stimulated by calcitriol. J Steroid Biochem Mol Biol. 2007; 103(3–5): 529–532. PubMed Abstract \| Publisher Full Text\n\nSun T, Zhao Y, Mangelsdorf DJ, et al.: Characterization of a region upstream of exon I.1 of the human CYP19 (aromatase) gene that mediates regulation by retinoids in human choriocarcinoma cells. Endocrinology. 1998; 139(4): 1684–1691. PubMed Abstract \| Publisher Full Text\n\nLou YR, Murtola T, Tuohimaa P: Regulation of aromatase and 5alpha-reductase by 25-hydroxyvitamin D(3), 1alpha,25-dihydroxyvitamin D(3), dexamethasone and progesterone in prostate cancer cells. J Steroid Biochem Mol Biol. 2005; 94(1–3): 151–157. PubMed Abstract \| Publisher Full Text\n\nTakayanagi R, Goto K, Suzuki S, et al.: Dehydroepiandrosterone (DHEA) as a possible source for estrogen formation in bone cells: correlation between bone mineral density and serum DHEA-sulfate concentration in postmenopausal women, and the presence of aromatase to be enhanced by 1,25-dihydroxyvitamin D3 in human osteoblasts. Mech Ageing Dev. 2002; 123(8): 1107–1114. PubMed Abstract \| Publisher Full Text\n\nTanaka S, Haji M, Takayanagi R, et al.: 1,25-Dihydroxyvitamin D3 enhances the enzymatic activity and expression of the messenger ribonucleic acid for aromatase cytochrome P450 synergistically with dexamethasone depending on the vitamin D receptor level in cultured human osteoblasts. Endocrinology. 1996; 137(5): 1860–1869. PubMed Abstract \| Publisher Full Text\n\nKinuta K, Tanaka H, Moriwake T, et al.: Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology. 2000; 141(4): 1317–1324. PubMed Abstract \| Publisher Full Text\n\nKrishnan AV, Swami S, Peng L, et al.: Tissue-selective regulation of aromatase expression by calcitriol: implications for breast cancer therapy. Endocrinology. 2010; 151(1): 32–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLundqvist J, Norlin M, Wikvall K: 1α,25-Dihydroxyvitamin D3 exerts tissue-specific effects on estrogen and androgen metabolism. Biochim Biophys Acta (BBA) - Molecular and Cell Biology of Lipids. 2011; 1811(4): 263–270. PubMed Abstract \| Publisher Full Text\n\nSwami S, Krishnan AV, Feldman D: 1alpha,25-Dihydroxyvitamin D3 down-regulates estrogen receptor abundance and suppresses estrogen actions in MCF-7 human breast cancer cells. Clin Cancer Res. 2000; 6(8): 3371–3379. PubMed Abstract\n\nSimboli-Campbell M, Narvaez CJ, VanWeelden K, et al.: Comparative effects of 1,25(OH)2D3 and EB1089 on cell cycle kinetics and apoptosis in MCF-7 breast cancer cells. Breast Cancer Res Treat. 1997; 42(1): 31–41. PubMed Abstract \| Publisher Full Text\n\nJames SY, Mackay AG, Binderup L, et al.: Effects of a new synthetic vitamin D analogue, EB1089, on the oestrogen-responsive growth of human breast cancer cells. J Endocrinol. 1994; 141(3): 555–563. PubMed Abstract \| Publisher Full Text\n\nStoica A, Saceda M, Fakhro A, et al.: Regulation of estrogen receptor-alpha; gene expression by 1, 25-dihydroxyvitamin D in MCF-7 cells. J Cell Biochem. 1999; 75(4): 640–651. PubMed Abstract\n\nLarsen SS, Heiberg I, Lykkesfeldt AE: Anti-oestrogen resistant human breast cancer cell lines are more sensitive towards treatment with the vitamin D analogue EB1089 than parent MCF-7 cells. Br J Cancer. 2001; 84(5): 686–690. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSwami S, Krishnan AV, Peng L, et al.: Transrepression of the estrogen receptor promoter by calcitriol in human breast cancer cells via two negative vitamin D response elements. Endocr Relat Cancer. 2013; 20(4): 565–577. PubMed Abstract \| Publisher Full Text\n\nHalloran BP, DeLuca HF: Effect of vitamin D deficiency on fertility and reproductive capacity in the female rat. J Nutr. 1980; 110(8): 1573–1580. PubMed Abstract\n\nKwiecinski GG, Petrie GI, DeLuca HF: Vitamin D is necessary for reproductive functions of the male rat. J Nutr. 1989; 119(5): 741–744. PubMed Abstract\n\nKrishnan AV, Swami S, Feldman D: Equivalent anticancer activities of dietary vitamin D and calcitriol in an animal model of breast cancer: importance of mammary CYP27B1 for treatment and prevention. J Steroid Biochem Mol Biol. 2013; 136: 289–295. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChristensen GL, Jepsen JS, Fog CK, et al.: Sequential versus combined treatment of human breast cancer cells with antiestrogens and the vitamin D analogue EB1089 and evaluation of predictive markers for vitamin D treatment. Breast Cancer Res Treat. 2004; 85(1): 53–63. PubMed Abstract \| Publisher Full Text\n\nKrishnan AV, Swami S, Feldman D: Vitamin D and breast cancer: inhibition of estrogen synthesis and signaling. J Steroid Biochem Mol Biol. 2010; 121(1–2): 343–348. PubMed Abstract \| Publisher Full Text\n\nWelsh J: Targets of vitamin D receptor signaling in the mammary gland. J Bone Miner Res. 2007; 22(Suppl 2): V86–V90. PubMed Abstract \| Publisher Full Text\n\nIngraham BA, Bragdon B, Nohe A: Molecular basis of the potential of vitamin D to prevent cancer. Curr Med Res Opin. 2008; 24(1): 139–149. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5396", "date": "02 Feb 2015", "name": "Daniel Bikle", "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 is comprised of three parts: first a review of the mechanisms of action of vitamin D, second a review of steroid hormone metabolism, and third a review of primarily the authors research into the role of 1,25(OH)2D in regulating steroid hormone metabolism. The first sections provide a brief but effective introduction to the authors main focus on his own work regarding the regulation by vitamin D of steroidogenesis. The major criticism I have is that this last section reveals how preliminary the evidence is that vitamin D has a physiologically significant role in steroidogenesis. The work has been done primarily in cancer cells in vitro. No compelling in vivo data are provided to show its physiologic importance, although animal models with deletions in VDR and CYP27B1 have been available for years. Nevertheless, this review opens up an interesting area for future research, and as such serves as a useful stimulus for that research to occur.", "responses": [] }, { "id": "8350", "date": "18 May 2015", "name": "Takashi Yazawa", "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\nLundqvist has overviewed the effects of Vitamin D on steroidogenesis. Overall, the manuscript is disorganized. In addition, there are many serious issues:The references are improper. There are too few for a review article. Additionally, the author references some articles which are already retracted. There are some unclear sentences. For example:Page 2, last line: \"For example, 1α,25-dihydroxyvitamin D3 increases the gene expression of CYP24A1 while it increases the gene expression of CYP27B1.\" Why are both genes are upregulated by vitamin D?Page 3, Adrenal steroidogenesis: “[...] and zona reticularis is the point of synthesis for adrenal androgens (e.g. DHEA)\" What is the point? The description of steroidogenesis is not sufficient. Abstract: “Steroid hormones are synthesized in steroidogenic tissues such as [...]” Why are breast and prostate tissues steroidogenic? They hardly perform de novo synthesis of steroid hormones, at least under normal conditions. Page 3, Adrenal steroidogenesis: “The adrenal steroidogenesis is quantitatively regulated by the transcription of CYP11A1 (cholesterol side-chain cleavage enzyme) and the activity of steroidogenic acute regulatory protein (StAR).\" The transcriptional regulation of StAR gene is also a rate-limiting step for steroidogenesis. P3, Adrenal steroidogenesis: “The qualitative regulation of adrenal steroidogenesis, determining which type of steroid that will be produced, is performed by the transcription and activity of CYP17A1.” This sentence is an overstatement. The expression of 3β-HSD is also important for zone-specific adrenal steroidogenesis.", "responses": [] } ]	1	https://f1000research.com/articles/3-155
https://f1000research.com/articles/3-153/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "GeneMANIA: Fast gene network construction and function prediction for Cytoscape", "authors": [ "Jason Montojo", "Khalid Zuberi", "Harold Rodriguez", "Gary D. Bader", "Quaid Morris", "Khalid Zuberi", "Harold Rodriguez", "Gary D. Bader", "Quaid Morris" ], "abstract": "The GeneMANIA Cytoscape app enables users to construct a composite gene-gene functional interaction network from a gene list. The resulting network includes the genes most related to the original list, and functional annotations from Gene Ontology. The edges are annotated with details about the publication or data source the interactions were derived from. The app leverages GeneMANIA’s database of 1800+ networks, containing over 500 million interactions spanning 8 organisms: A. thaliana, C. elegans, D. melanogaster, D. rerio, H. sapiens, M. musculus, R. norvegicus, and S. cerevisiae. Users may also import their own organisms, networks, and expression profiles. The app is compatible with Cytoscape versions 2 and 3.", "keywords": [ "The GeneMANIA Cytoscape1 app enables users to construct a weighted composite functional interaction network from a list of genes. Each node represents a gene and its products. The app uses the GeneMANIA algorithm2 to find other genes and gene products that are most related to the original list", "and shows how they are related." ], "content": "Introduction\n\nThe GeneMANIA Cytoscape1 app enables users to construct a weighted composite functional interaction network from a list of genes. Each node represents a gene and its products. The app uses the GeneMANIA algorithm2 to find other genes and gene products that are most related to the original list, and shows how they are related.\n\nThe app provides access to most of the features of the GeneMANIA prediction server3 while removing limitations on gene list length, and the maximum size of the resulting network. The app also allows predictions to be made on user-defined organisms and arbitrarily large custom networks.\n\nGeneMANIA uses a database of organism-specific weighted networks to construct the resulting composite network. The database includes over 1800 networks, containing over 500 million interactions for 8 organisms: A. thaliana, C. elegans, D. melanogaster, D. rerio, H. sapiens, M. musculus, R. norvegicus, and S. cerevisiae. The networks are organized into groups such as co-expression, where edges are derived from expression profiles, and shared protein domains, where edges represent genes that encode proteins with similar domains. Users may select any combination of these as the basis of the composite network they construct for their gene list.\n\nPrior to construction, the selected networks are each assigned a weight by the GeneMANIA algorithm. The weight of each edge is multiplied by the weight of the containing network. Next, the union of all edges in the network is taken. In the case of multiple edges between any pair of nodes, the edges are collapsed into one and assigned a weight equal to the sum of the individual edge weights. The query genes are assigned a label value of 1, while all other genes are 0. Label propagation is then applied to the entire network2 and the resulting labels are saved as the score attribute in the node table. This score indicates the relevance of each gene to the original list based on the selected networks. Higher scores indicate genes that are more likely to be functionally related. Users may extend their original gene list by adding these top ranking genes to their network. They can also choose not to add any other genes so they can visualize how the members of their list are connected.\n\nInstead of providing the user with the composite network used during label propagation, the Cytoscape app displays at most one edge for each type of network that contributed to the gene scores (Figure 1). For example, if five co-expression networks and two physical interaction networks contained an edge between the same pair of genes, the resulting network would contain one co-expression edge and one physical interaction edge for that pair. The edges are annotated with the original edge weights, the source networks from which those weights originate, relevant publications, and details about how the data was collected or processed (Figure 2). The nodes are annotated with Gene Ontology4 terms, alternate identifiers and synonyms.\n\nThe circles are genes and the diamonds are protein domain attributes. Up to 20 most related genes and 20 most related attributes are shown. The red genes are annotated with DNA repair, as indicated in the Functions tab.\n\nThe source networks are grouped by type (e.g. co-expression) and list each network weight, as well as the sum of the weights of the networks in each group. Citations and links to relevant publications and data sources and provided where possible.\n\n\nImplementation\n\nThe GeneMANIA app is an update to the GeneMANIA plugin for Cytoscape 25. The app preserves runtime compatibility with older versions of Cytoscape. It is distributed as a universal binary that runs on every release of Cytoscape since version 2.6.3. Figure 3 illustrates how we architected the software to enable the same code to run in multiple environments. The GeneMANIA Engine module, which implements the algorithm, is an independent layer that is also used directly by the GeneMANIA prediction web server. The App Core module includes highly parallelized command line tools for function prediction and cross validation6 on multiprocessor clusters and multicore workstations. It also contains an abstraction layer to provide access to a small subset of Cytoscape’s functionality through high-level Application Programming Interface (API). This alternative API effectively decouples the app implementation from a particular version of Cytoscape, allowing the same code to drive a Cytoscape 2 plugin and Cytoscape 3 app.\n\nThe user-provided gene list is used to select the most relevant interactions from the GeneMANIA database. The resulting network is visualized in Cytoscape.\n\nThe app provides access to all previous editions of the GeneMANIA database dating back to the initial September 23, 2010 release. New data updates will also be supported as they become available. As of the March 3, 2011 database release, two subsets of the data are available for users with special requirements. The core subset is roughly 20% of the size of the full database and only includes networks that are selected by default3. The open license subset only includes network data with no restrictions on use. For example, networks derived from I2D7 and HPRD8 are excluded from this subset since their standard licenses prohibit commercial use of their data.\n\nThe networks are stored on disk as compact binary sparse matricies, which are used directly by GeneMANIA’s network integrator. This representation allows networks to be loaded quickly and used immediately without transformation into a different data structure. Gene and network metadata, including descriptions and provenance details, are stored in a Lucene index. This allows fast retrieval of metadata and gene name autocompletion as users type in their list.\n\nUnlike the GeneMANIA prediction server which only supports 8 organisms, the app allows users to perform predictions on their own organisms. To import an organism into a user’s local database, the user needs to provide a tab-delimited file containing the organism’s genome, where each row contains the primary identifier of a gene followed by alternate identifiers and synonyms. From there, users may import tab-delimited network data or expression profiles. Users may also import networks or expression profiles they have loaded into Cytoscape. The app can also be used with non-biological data such as social networks, where the nodes are individuals and edges represent various relationships between them.\n\n\nResults\n\nTo demonstrate the steps involved with performing predictions on custom organisms not already provided by GeneMANIA, Ensembl9 Gene IDs and their associated gene names for Felis catus were imported from BioMart10 and imported into GeneMANIA as an organism. Data set GSE46431 was downloaded from the Gene Expression Omnibus (GEO)11 and imported directly as expression profile data to yield a coexpression network. On a 2.3 GHz Intel Core i7 3615QM system with 16 GB RAM and SSD storage, it took approximately 5 minutes to import the data. Using this network, the app was used to find and display the 20 genes most related to ASIP, which took 1 second (Figure 4).\n\nThe expression profiles from this dataset were converted into a co-expression network using the GeneMANIA app.\n\n\nConclusions\n\nThe GeneMANIA app extends the capabilities of the GeneMANIA prediction server by allowing users to quickly construct networks from gene lists for custom organisms and network data without imposing any limits on the size of the inputs or output while retaining provenance of the source data. The app also allows users to replicate past results by providing access to all publicly-released GeneMANIA datasets.\n\n\nSoftware availability\n\nSoftware available from the Cytoscape’s App Manager or the App Store: http://apps.cytoscape.org/apps/GeneMania.\n\nLatest source code: https://github.com/GeneMANIA/genemania.\n\nSource code as at the time of publication: https://github.com/F1000Research/genemania/releases/tag/V1.0\n\nArchived source code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.1052312\n\nLicense: LGPL 2.1: https://www.gnu.org/licenses/old-licenses/lgpl-2.1.html", "appendix": "Author contributions\n\n\n\nJM wrote the manuscript. JM and KZ wrote the software. JM, KZ, and HR prepared the prepackaged network data. GDB and QM designed and supervised the project.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by the Ontario Government’s Ministry of Research and Innovation via the Global Leadership Round in Genomics & Life Sciences (GL2) program (assigned to GDB and QM); and the National Resource for Network Biology (U.S. National Institutes of Health, National Center for Research Resources grant number P41GM103504, assigned to GDB).\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\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\nMostafavi S, Morris Q: Fast integration of heterogeneous data sources for predicting gene function with limited annotation. Bioinformatics. 2010; 26(14): 1759–1765. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZuberi K, Franz M, Rodriguez H, et al.: GeneMANIA prediction server 2013 update. Nucleic Acids Res. 2013; 41(Web Server issue): W115–W122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris MA, Clark J, Ireland A, et al.: Gene Ontology Consortium. The Gene Ontology (GO) database and informatics resource. Nucleic Acids Res. 2004; 32(Database issue): D258–D261. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontojo J, Zuberi K, Rodriguez H, et al.: GeneMANIA Cytoscape plugin: fast gene function predictions on the desktop. Bioinformatics. 2010; 26(22): 2927–2928. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontojo J, Zuberi K, Shao Q, et al.: Network Assessor: an automated method for quantitative assessment of a network’s potential for gene function prediction. Front Genet. 2014; 5: 123. 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\nKeshava Prasad TS, Goel R, Kandasamy K, et al.: Human Protein Reference Database--2009 update. Nucleic Acids Res. 2009; 37(Database issue): D767–D772. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFlicek P, Amode MR, Barrell D, et al.: Ensembl 2014. Nucleic Acids Res. 2014; 42(Database issue): D749–D755. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKinsella RJ, Kähäri A, Haider S, et al.: Ensembl Biomarts: a hub for data retrieval across taxonomic space. Database (Oxford). 2011; 2011: bar030. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarrett T, Wilhite SE, Ledoux P, et al.: NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013; 41(Database issue): D991–D995. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontojo J, Zuberi K, Rodriguez H, et al.: F1000Research/genemania. ZENODO. 2014. Data Source" }	[ { "id": "5285", "date": "21 Jul 2014", "name": "Giovanni Scardoni", "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 paper the authors present the GeneMANIA Cytoscape app. It allows users to construct gene networks from a gene list. The paper is well written and complete.Minor comment:The database contains over 500 million interactions spanning 8 organisms: A. thaliana, C. elegans, D. melanogaster, D. rerio, H. sapiens, M. musculus, R. norvegicus, and S. cerevisiae. It is not clear how this interactions are obtained (literature, other database ...) More details about it would improve the paper.", "responses": [] }, { "id": "5567", "date": "01 Aug 2014", "name": "Rosalba Giugno", "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 updated version of an affirmed work in the literature. A few comments follow. The authors should highlight the upgrading aspects with respect to the previous work. Moreover, it would be useful to report the used database sources, the biggest retrieved network and how they deal with such a network with respect to the ‘unlimited’ list of genes (visualization and running time).", "responses": [] } ]	1	https://f1000research.com/articles/3-153
https://f1000research.com/articles/3-150/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "Porting and using PanGIA for Cytoscape 3: challenges and solutions", "authors": [ "David Welker", "Barry Demchak", "Barry Demchak" ], "abstract": "Much of the biologically significant functionality in Cytoscape is contained within third-party add-ons, called plugins in Cytoscape 2 and apps in Cytoscape 3. In the transition from Cytoscape 2 to Cystoscape 3, some of the underlying assumptions upon which plugins relied changed, requiring a significant porting effort for plugins to work as Cytoscape 3 apps. PanGIA is a Cytoscape add-on (http://apps.cytoscape.org/apps/pangia) designed to analyze and visualize genetic interaction data in light of physical interaction data. In order to convert the PanGIA plugin to an app, various challenges, including those related to a transformed data model, concurrency, and randomization had to be overcome. In the process, the ability to control randomization was added to the GUI, a feature which was not only integral to the porting process, but which also ensures more easily reproducible scientific analysis for PanGIA users. Most authors of Cytoscape 2 plugins will face similar challenges porting their software to work with Cytoscape 3, and this paper gives details of how the PanGIA port addressed them.", "keywords": [ "PanGIA (which stands for Physical and Genetic Interaction Alignment) is a Cytoscape plugin originally developed for Cytoscape 2", "and which has now been released for Cytoscape 3. The purpose of PanGIA is to help determine the functional significance of genetic interaction data", "which is being increasingly uncovered by a plethora of modern high-throughput technologies. It identifies genetic interactions in the context of physical protein-protein interactions by identifying both “within-cluster” physical and genetic interactions", "which may signify either functional protein complexes or signaling pathways", "and “between-cluster” genetic interactions", "which often identify synergistic or compensatory relationships between clusters Srivas et al.1. In order to ensure the ability to perform such analysis was widely available to scientists", "PanGIA was originally released as a novel open source software tool that worked as a plug-in for Cytoscape 2. Due to sufficient demand by researchers", "we have ported PanGIA as an app (the new term for plugin) for Cytoscape 3." ], "content": "Introduction\n\nPanGIA (which stands for Physical and Genetic Interaction Alignment) is a Cytoscape plugin originally developed for Cytoscape 2, and which has now been released for Cytoscape 3. The purpose of PanGIA is to help determine the functional significance of genetic interaction data, which is being increasingly uncovered by a plethora of modern high-throughput technologies. It identifies genetic interactions in the context of physical protein-protein interactions by identifying both “within-cluster” physical and genetic interactions, which may signify either functional protein complexes or signaling pathways, and “between-cluster” genetic interactions, which often identify synergistic or compensatory relationships between clusters Srivas et al.1. In order to ensure the ability to perform such analysis was widely available to scientists, PanGIA was originally released as a novel open source software tool that worked as a plug-in for Cytoscape 2. Due to sufficient demand by researchers, we have ported PanGIA as an app (the new term for plugin) for Cytoscape 3.\n\nThe challenges we encountered in the porting process and our solutions to them are the primary focus of this paper. As a secondary focus, this paper provides a brief overview of how to use PanGIA and the results it produces. Readers of this paper who are not app developers interested in porting their plugin from Cytoscape 2 to Cytoscape 3 may want to skip the “Porting PanGIA” section of the paper. The remainder of this paper consists of three primary parts: a porting section that highlights the main technical challenges we encountered in porting PanGIA, a brief overview of how to run PanGIA, and a final section that provides a brief overview of the meaning of PanGIA’s outputs.\n\n\nPorting PanGIA\n\nTo port a Cytoscape 2 plugin to a Cytoscape 3 app, a good understanding of Java and Swing is necessary. Often, deficiencies in knowledge of these technologies can be overcome by referencing Oracle’s Java Tutorials (http://docs.oracle.com/javase/tutorial/). The rest of this section assumes basic familiarity with Java and Swing. Additionally, for those looking for basic information on how to port an app to Cytoscape 3, a good place to start is the Cytoscape Plugin Porting Guide (http://wiki.cytoscape.org/Cytoscape_3/AppDeveloper/PluginPortingGuide).\n\nIn our port of PanGIA, we encountered three main challenges. First, the manner in which data is stored concerning nodes, edges, and networks is significantly different in Cytoscape 2 versus Cytoscape 3, and it was rather time-consuming to perform the conversion. Second, it was necessary to explicitly ensure that the Swing dialog messages that PanGIA displayed to the users were run on the EDT in order to avoid deadlocks (Whereas in Cytoscape 2, tasks often run on Java’s event dispatch thread (“EDT”) by default, in Cytoscape 3, they do not.) Finally, since PanGIA involves heuristic algorithms with a random element, it was necessary to devise a means to control such randomization so that the same results on the same data could be achieved in both Cytoscape 2 and Cytoscape 3 in order to confirm correctness. Besides these three main challenges, there were numerous smaller challenges. For example, we had to create a CyActivator to register objects relating to the new Cytoscape OSGi framework, and to create context menu items.\n\nA useful reference for performing OSGi-related developer work is the Cytoscape Samples Repository (https://github.com/cytoscape/cytoscape-samples), since it contains many hands-on examples of such work (such as all the steps for registering context menu items) that are not documented as well elsewhere. The Cytoscape 3 App Cookbook (http://wiki.cytoscape.org/Cytoscape_3/AppDeveloper/Cytoscape_3_App_Cookbook) is also an invaluable resource that ought to be consulted regularly by developers looking to port their plugins.\n\nThe remainder of this section briefly describes each of the three main challenges to porting that we encountered in turn.\n\nThe most time-consuming aspect of porting PanGIA was adapting to the numerous changes in the manner in which data is stored in Cytoscape 3 compared to Cytoscape 2. While this type of work is usually straight-forward, depending on the plugin, changes are likely to be extensive since most plugins interact with the underlying network data model extensively.\n\nOne issue with the Cytoscape 3 data model is likely to be counterintuitive to novice plugin porters in particular. In general, when adding data to a node table or edge table, one usually must first add an appropriate column to the table object. Based on this, one might intuitively think it is most common to put the data directly into the table object as well. But such an inference would be incorrect; it is much more common and convenient to add the data using the network object. This is done because before new data can be added to the table, a new row must be created. But to create a new row on the table object, one must call an appropriate method on the network object, not the table object. After doing this, although it is possible to add the data using the table object directly, it is much more common to use a convenience method on the network object to add the data to the table indirectly. The code snippet below is an example of working with the data model in Cytoscape 3 in the context of a node table:\n\n\n\nThis contrasts with Cytoscape 2, where data was usually added to CyAttributes (the approximate equivalent of CyTable) directly, rather than most commonly going through another object. The code snippet below shows the typical case of data being added directly to the CyAttributes object directly:\n\n\n\nUsing one object to create a column and then another to create a row and add data is just an example of one of the more tricky transformation issue that developers are likely to face. In general, Cytoscape’s powerful and flexible data model is not always intuitive at first, but the process becomes easier with time. To port a plugin to Cytoscape 3, the first task is usually understanding the Cytoscape 3 data model and performing the conversion.\n\nPart of the standard mechanism in Java for handling concurrency is to ensure that nearly all invocations of Swing components occur on the EDT. (See The Java Tutorial: The Event Dispatch Thread at http://docs.oracle.com/javase/tutorial/uiswing/concurrency/dispatch.html). In Cytoscape 2, tasks were invoked on the current thread, which in many cases was the EDT. This allowed some plugins to implicitly rely on the fact that they were probably being executed on the EDT. However, in Cytoscape 3, every task is spawned on a new thread, which means that when work is done in a task, by default, it does not run on the EDT. Because PanGIA relied on the old behavior, it was necessary to explicitly make sure that any messages to be displayed to the user using Swing dialogs were run on the EDT. This was typically done using the SwingUtilities.invokeLater method using an anonymous inner class. Code that needs to be run on the EDT, such as any invocations of Swing components, is simply executed in the run method of an anonymous Runnable object which is passed as a parameter into the invokeLater method. The code snippet below illustrates:\n\n\n\nSince this is an anonymous inner class, if it uses any data from the method or class it is embedded in, that data must be declared final. In the instances where this is a problem, the goal of data access can usually be accomplished by declaring a new final variable and assigning the non-final data to that. Then, it is the new final variable, not the original data, that would be accessed in the run method. Finally, although it is permissible to do non-Swing work on the EDT, such work ought to be minimized, as this has the potential to cause the GUI to hang because such work is given a higher priority than repainting the GUI.\n\nIn PanGIA, the algorithms used to determine the clusters of genes (or modules, as PanGIA refers to them) use a random number generator that influences both the assignment of genes to modules and the number of modules produced. Therefore, this random number generator made it very difficult to determine the correctness of the port of PanGIA from Cytoscape 2 to Cytoscape 3, since different clustering arrangements occurred simply as a matter of chance. Fortunately, the random numbers in Java and other programming languages are not truly random, but rather pseudorandom. (See the API for the Random class at http://docs.oracle.com/javase/7/docs/api/java/util/Random.html). This means that, unlike with truly random numbers, we can eliminate unpredictability and instead create repeatability. When unpredictability isn’t desired, it is possible to take control of the random number generation process by setting an initial value (called a seed) for the Random class explicitly. For our port of PanGIA, we took advantage of this to prevent unpredictability in order to verify that the same results were produced by both versions of the PanGIA add-on.\n\nPanGIA for Cytoscape 2 had multiple locations where instances of the Random class were instantiated in a manner that did not control randomness. In order to control randomness for these instances, we created a single seed variable and added the ability to set this seed variable from the GUI. We then changed each of these locations so that they constructed Random objects using this seed variable. Upon choosing any arbitrary value for the seed, we saw that results were now reproducible, just so long as the same arbitrary value was consistently chosen. In order to compare results across Cytoscape 2 and Cytoscape 3 to ensure correctness, we also added the same seed functionality to the PanGIA plugin for Cytoscape 2 and saw that when the same data was used in both versions, the same outputs were produced. This sort of black box testing was only possible when additional functionality was added to control randomization.\n\n\nRunning PanGIA\n\nRunning PanGIA is fairly straightforward. First, one must load the physical and genetic network. Then one must set appropriate parameters using PanGIA’s graphical user interface (GUI). Finally, one hits the Search button and PanGIA runs for an amount of time that depends on the size of the physical and genetic networks and outputs the results. The meaning of the input parameters and advice on setting them are well explained in Srivas et al.1. With only small deviations, the very detailed steps in that paper are still applicable and we recommend that users who need a detailed understanding of how to run PanGIA continue to reference that paper as their primary guide to setting input parameters.\n\nHowever, there is one new feature that PanGIA users will find interesting which does not exist in the version of PanGIA released to work with Cytoscape 2, and that is the ability to control randomization by setting the seed parameter in the GUI. Users should know that PanGIA’s core algorithms are non-deterministic because of their dependence on random number generation. This means that with the previous version of PanGIA, a user’s results may not be perfectly reproducible in the future. But if the user needs reproducibility for their research, with PanGIA for Cytoscape 3, it is now possible. In the new PanGIA GUI, there is a new parameter that can be set called Seed. The user simply sets this to any integer of their choosing (in the range of −263 to 263−1). As long as the same number is used each time PanGIA is run on the same data set with all the other parameters held constant, the same results will be output.\n\n\nVisualizing and analyzing results\n\nAfter running PanGIA, a large number of networks will result. One of these networks will be the overview network while all the others will consist of modules of genes with significant physical protein-protein interactions and associated genetic interactions. These networks will allow the user to visualize the overview network (Figure 1), within-cluster interactions (Figure 2), and between-cluster interactions (Figure 3). The remainder of this section will consider each of these in turn.\n\nThe overview network consists of a higher level view of the results, where the nodes are the modules that PanGIA found and the edges represent genetic interactions that exist between those modules. From the overview network, the user can see between-cluster genetic interaction generally, but not the specific genes in each module that are involved. Thus, the overview network is good for a high-level overview of between-cluster genetic interactions between modules as well as a sense of how many and which modules were discovered.\n\nWithin-cluster interactions can be examined by selecting a node within the overview network and right-clicking and choosing Apps \| PanGIA \| Create Detailed View. Black edges represent physical interactions, light yellow edges represent positive genetic interactions, and light blue edges represent negative genetic interactions.\n\nTo visualize between-cluster interactions, the user selects two or more nodes in the overview network and right-clicks to bring up the context menu and chooses Apps \| PanGIA \| Create Detailed View. For the selected nodes, the user will still see within-cluster physical and genetic interactions. In addition, the user will see the genetic interactions that connect the two or more modules and the specific genes involved.\n\n\nConclusions\n\nIn this paper, we have explored some of the challenges we faced when porting PanGIA which may be applicable to developers looking to port other plugins. We also briefly explained how running PanGIA may be affected by the new random seed feature and briefly examined the outputs from running PanGIA.\n\nSoftware available from http://apps.cytoscape.org/apps/pangia\n\nLatest source code https://github.com/idekerlab/pangia3/\n\nSource code as at the time of publication https://github.com/F1000Research/pangia3/releases/tag/V1.0\n\nArchived source code as at the time of publication http://www.dx.doi.org/105281/zenodo.104592\n\nLicense Lesser GNU Public License v2.1", "appendix": "Author contributions\n\n\n\nDW ported PanGIA and wrote this article. BD supervised the port and reviewed this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported with funding from the National Resource for Network Biology (NRNB) under award numbers P41 RR031228 and GM103504 assigned to Trey Ideker.\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 Dr. Dieter Wolf and Dr. Rohith Srivas for reviewing our port of PanGIA to confirm its correctness and making other suggestions.\n\n\nSupplementary data\n\nThe data necessary to reproduce the figures used in this paper were taken from the website for Srivas et al.1 (http://prosecco.ucsd.edu/PanGIA/) (available as physical_network.txt, genetic_network.txt, and annotations.txt). Please load the physical and genetic networks before the annotations.\n\n\nReferences\n\nSrivas R, Hannum G, Ruscheinski J, et al.: Assembling global maps of cellular function through integrative analysis of physical and genetic networks. Nat Protoc. 2011; 6(9): 1308–1323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWelker D, Demchak B: F1000Research/Pangia3. ZENODO. 2014. Data Source" }	[ { "id": "5294", "date": "14 Jul 2014", "name": "Brian Kidd", "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 provides useful technical advice for Cytoscape app developers looking to port plugins from version 2 to apps in version 3. The technical advice is provided in the context of a specific application PanGIA, which helps determine the functional significance of genetic interaction data. The case study of PanGIA describes:How to deal with the data model and GUI event handling changes for developing software within the Cytoscape version 3 framework. What upgrades have been introduced in the port from v2 to v3, and some expected results when running the software.Overall the authors do a nice job of giving development advice and technical details for creating a Cytoscape 3 app if one is familiar with creating plugins for Cytoscape 2. In addition, the authors describe the new functionality that was added to the PanGIA app (random seed for obtaining reproducible results from the heuristic algorithm) and offer a small tutorial for anyone wanting to use the app for their research.This article is clearly written and provides a helpful resource for Cytoscape developers as well as end-users.", "responses": [] }, { "id": "5296", "date": "14 Jul 2014", "name": "Alfredo Pulvirenti", "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 paper the authors describe their experience in the porting of PanGIA to Cytoscape 3. The paper aims to: (i) point out to the developers the main tricks and challenges to perform the porting from Cytoscape 2 to 3; (ii) describe the new feature in the last PanGIA plugin.The paper is interesting and it certainly could result in being useful to developers and users. Minor issue: The authors emphasize the novel feature allowing reproducing the results by controlling the seed of the random number generator. This is certainly an important point, however I would reduce the paragraph since it describes really basic programming stuff. On the other hand, I would stress the point that such a “seed control” should be always available in all the systems using randomization. Furthermore, in the “running PanGIA paragraph” the description of the “seed setting” is quite repetitive and should be also reduced.", "responses": [] }, { "id": "5297", "date": "21 Jul 2014", "name": "Christian Lopes", "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 these technical challenges, faced by the authors when porting the PanGIA plugin from Cytoscape 2 to Cytoscape 3:The new data model is very different from the one in Cytoscape 2. Tasks that need to display messages on the GUI are not guaranteed to run on the EDT on Cytoscape 3, so the developer has now to explicitly ensure that in order to avoid deadlocks. The authors had to add a new randomization control feature to the old and the new version of PanGIA, so the output of both versions can produce the same results.The paper also provides overviews of how to run PanGIA and the meaning of the networks it produces.The article is consistent, well written and provides useful technical information for developers who want to port a Cytoscape 2 plugin to a Cytoscape app.I have only one minor concern:I totally agree that \"Cytoscape’s powerful and flexible data model is not always intuitive at first”, but the code snippet that exemplifies the use of the new data model is a bit longer than necessary, especially if the main goal is to compare it with the other code snippet, which shows how Cytoscape 2 data model is typically used. It would be fairer to use an example that only deals with the data model, and not network or node creation, because: (a) the Cytoscape 2 code snippet does not show network/node creation; (b) network/node creation is not necessary every time data is added, so the given example looks more complicated than it actually is. I think a more equivalent example would be something like this:public <T> void addDataToTable(CyNetwork network, CyNode node, Class<T> type, T data) {\n\n//Get node table.\n\nCyTable table = network.getDefaultNodeTable();\n\n//Create the column in the table object.\n\nif (table.getColumn(\"MyColumn\") == null)\n\ntable.createColumn(\"MyColumn\", type, false);\n\n//Most commonly, data for the node is\n\n//added using the network object, not the table object.\n\nnetwork.getRow(node).set(\"MyColumn\", data);}", "responses": [] } ]	1	https://f1000research.com/articles/3-150
https://f1000research.com/articles/3-148/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "BridgeDb app: unifying identifier mapping services for Cytoscape", "authors": [ "Jianjiong Gao", "Chao Zhang", "Martijn van Iersel", "Li Zhang", "Dong Xu", "Nikolaus Schultz", "Alexander R. Pico", "Chao Zhang", "Martijn van Iersel", "Li Zhang", "Dong Xu", "Nikolaus Schultz", "Alexander R. Pico" ], "abstract": "The BridgeDb app for Cytoscape allows users to map and annotate identifiers of genes, proteins and metabolites in the context of biological networks. The app greatly simplifies the identifier mapping process in Cytoscape by providing a unified interface to different mapping resources and services. The app also provides a programming interface via Cytoscape Commands that can be utilized for identifier mapping by other Cytoscape apps. In this article we provide a technical guide to the BridgeDb app for mapping identifiers in Cytoscape.", "keywords": [ "Cytoscape1 is a powerful network visualization tool and platform for data integration and analysis. However", "identifier mapping remains a challenge when working with biological data from different sources. We developed the BridgeDb app for Cytoscape to provide utilities for mapping and annotating identifiers in the network context. Built on the BridgeDb open-source framework for identifier mapping2", "the BridgeDb app provides a graphical user interface (GUI) for users and a command interface for other Cytoscape apps to perform identifier mapping in Cytoscape networks. As a result", "BridgeDb enables or simplifies the processes of identifier translation", "biological entity unification", "and functional annotation." ], "content": "Introduction\n\nCytoscape1 is a powerful network visualization tool and platform for data integration and analysis. However, identifier mapping remains a challenge when working with biological data from different sources. We developed the BridgeDb app for Cytoscape to provide utilities for mapping and annotating identifiers in the network context. Built on the BridgeDb open-source framework for identifier mapping2, the BridgeDb app provides a graphical user interface (GUI) for users and a command interface for other Cytoscape apps to perform identifier mapping in Cytoscape networks. As a result, BridgeDb enables or simplifies the processes of identifier translation, biological entity unification, and functional annotation.\n\n\nImplementation\n\nThe BridgeDb identifier mapping framework (http://bridgedb.org/)2 was designed to provide standardized access to gene, protein and metabolite identifier mapping services such as Ensembl BioMart3, Synergizer4, PICR5 and BridgeDb web services and BridgeDb database files. By creating a Java-based abstract layer, BridgeDb enables bioinformatics applications to connect to different mapping resources through the same interface, which greatly alleviates the burden of exploring, maintaining and switching between resources.\n\nBuilt upon BridgeDb framework and API, the BridgeDb app for Cytoscape can be used to connect to different mapping resources and map identifiers in Cytoscape netwoks. The BridgeDb app was implemented based on the Cytoscape 3 API. Its predecessor was the CyThesarus plugin for Cytoscape 2. To take advantage of the new OSGi based architecture in Cytoscape 3 and its clearly defined and simplified API, we have rewritten the CyThesaurus plugin into an OSGi bundle app. Figure 1 illustrates the implementation details. The identifier mapping API of BridgeDb framework was wrapped by Cytoscape Task and TaskFactory API, which provide identifier mapping utilities to users through graphical user interfaces for managing mapping resources and performing identifier mapping. The BridgeDb app Tasks were also registered to Cytoscape as command services allowing other apps, such as Mosaic6 and NOA7 apps, and the Merge Network tool, to take advantage of BridgeDb app’s identifier mapping capacities.\n\n\nResults\n\nThe BridgeDb app provides an intuitive GUI to perform identifier mapping in Cytoscape networks, consisting of two interactive dialogs for resource management and identifier mapping, respectively, which can be opened via the menu item Apps/BridgeDb.\n\nThe resource management dialog allows user to add, remove, and select or deselect mapping resources. Once the resources are configured, they will be saved in a global property file and therefore shared among different Cytoscape sessions. Currently BridgeDb app supports local and remote delimited text files, BridgeDb database files, BridgeDb web service, BioMart web service, and Synergizer web service. Web services are easy to access and up to date and therefore are preferable for annotating small to medium-size networks (less than 1,000 nodes). It is highly recommended to download and use the BridgeDb database files for mapping identifiers in large networks in order to save time. Local delimited text files are useful when mapping between non-standard or customized identifiers. To avoid ambiguity, it is recommended to select only one resource unless multiple resources are believed to be complementary to each other. Particularly, resources for different species should not be selected at the same time. Database or text files are also preferable when reproducibility is essential. We will update the app to support more mapping resources as BridgeDb API keeps being developed.\n\nThe identifier mapping dialog is the main interface for mapping identifiers stored in the node table in the selected Cytoscape network. The user needs to choose the source identifier types, columns in the node table that contain the source identifiers, target identifier types, and columns to save the target identifiers. When mapping, all specified source identifier types in all selected resources will be queried for matching identifiers. If one source identifier can be mapped onto multiple target identifiers, all target identifiers can be saved as a list in the node table. If a target column exists in the node table, all values in the column will be overwritten with the target identifiers; otherwise, a new column will be created and filled with the target identifiers. Besides a set of supported identifier types (e.g., Entrez Gene ID and UniProt accession), a mapping resource such as BridgeDb database may also have a set of supported attributes (e.g., gene symbol and description).\n\nThe BridgeDb app also provides a set of Cytoscape commands, which can be used by scripting or by other Cytoscape apps (client apps) to take advantage of the identifier mapping capacities provided by BridgeDb app. Table 1 provides a selected list of commands supported by BridgeDb. Every command has an optional argument ‘appName’ in order to allow every client app to have its own set of identifier mapping resources. Morris et al. (setsApp8, published together in the Cytoscape App collection9) provided an example of how to use commands in client apps.\n\nArguments with asterisks (*) are required.\n\nUse cases of BridgeDb app include identifier translation, biological entity unification, and functional annotation in Cytoscape networks. Box 1 provides an example of how to use BridgeDb app to facilitate annotation and integration of networks from public databases. In particular, this example shows the steps to generate a TP53 interaction network by merging networks from different sources using BridgeDb app and Merge Network tool. Figure 2 illustrates the process of the Cytoscape app NOA7 annotating Cytoscape networks with Gene Ontology by utilizing BridgeDb app’s command services.\n\n1. Install and start Cytoscape 3.1.0 or above\n\n2. Select File > Import > Network > Public Databases. . .\n\n3. Enter search condition TP53 AND human, press Search, select database IntAct, and click Import\n\n4. After importing, click No when asking about whether to manually merge networks, and close the import dialog\n\n5. If needed, install BridgeDb from Apps > App Manager\n\n6. Select File > Import > Network > URL. . . > Example:Human Protein-Protein: Rual et al. . . . Nature 2005\n\n7. Use default when asking about setting of the new network\n\n8. Select Apps > BridgeDb > Manager ID Mapping Resources\n\n9. Click Databases, Select database type .bridge\n\n10. Click Download, download Hs_Derby_[date].zip, and unzip the file\n\n11. In Cytoscape, select the unzipped .bridge file\n\n12. Review supported identifier types and click Close\n\n13. Select network rual.sif in Network panel\n\n14. Select Apps > BridgeDb > Map Identifiers\n\n15. Select name as the Source Column in Node Table\n\n16. Select Entrez Gene as Source ID Type(s)\n\n17. Select Uniprot/TrEMBL as Target ID Type\n\n18. Click Insert in the destination ID types table\n\n19. Select Attribute: Symbol as the second Target ID Type\n\n20. Click OK, wait, then click No to close\n\n21. Search for TP53 in the Cytoscape search box\n\n22. Select Select > Nodes > First Neighbors of Selected Nodes > Undirected\n\n23. Select File > New > Network > From selected nodes, all edges\n\n24. Select Tool > Merge > Networks. . .\n\n25. Select the network from IntAct and rual.sif(1)\n\n26. Click Advanced Network Merge\n\n27. Select Matching columns: uniprotkb_accession for IntAct, and Uniprot/TrEMBL for rual.sif(1)\n\n28. Click Merge to get a TP53 interaction network merged from the two sources\n\n\nConclusions\n\nIn this paper, we presented the BridgeDb app for Cytoscape. By providing a unified interface to various mapping resources, BridgeDb app enables identifier mapping in the network context and therefore greatly eases the process of data integration in Cytoscape.\n\nSoftware available from: http://apps.cytoscape.org/apps/bridgedb\n\nLatest source code: https://github.com/jjgao/bridgedb.cytoscape\n\nSource code as at the time of publication: https://github.com/F1000Research/bridgedb.cytoscape\n\nArchived source code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.1046510\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\nJG, CZ, MI and ARP participated in the design of the described software. JG and MI participated in the development of BridgeDb framework. JG, CZ and LZ implemented the BridgeDb app. JG, CZ, MI, DX, NS and ARP contributed to the development and writing of this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was partially funded by the Google Summer of Code program (JG, CZ, and LZ), the Robertson Foundation (NS), and by NIGMS R01-GM100039 (ARP).\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 Scooter Morris for technical support on command interface, Kristina Hanspers for beta testing, and Gary Bader, Mike Smoot, Keiichiro Ono and David States for providing useful comments.\n\n\nReferences\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\nvan Iersel MP, Pico AR, Kelder T, et al.: The BridgeDb framework: standardized access to gene, protein and metabolite identifier mapping services. BMC bioinformatics. 2010; 11(1): 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKasprzyk A, Keefe D, Smedley D, et al.: EnsMart: a generic system for fast and flexible access to biological data. Genome Res. 2004; 14(1): 160–169. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerriz GF, Roth FP: The Synergizer service for translating gene, protein and other biological identifiers. Bioinformatics. 2008; 24(19): 2272–2273. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCôté RG, Jones P, Martens L, et al.: The Protein Identifier Cross-Referencing (PICR) service: reconciling protein identifiers across multiple source databases. BMC bioinformatics. 2007; 8(1): 401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang C, Hanspers K, Kuchinsky A, et al.: Mosaic: making biological sense of complex networks. Bioinformatics. 2012; 28(14): 1943–1944. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang C, Wang J, Hanspers K, et al.: NOA: a cytoscape plugin for network ontology analysis. Bioinformatics. 2013; 29(16): 2066–2067. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorris JH, Wu A, Doncheva NT, et al.: setsapp: Set operations for cytoscape nodes and edges. Manuscript submitted for publication. F1000Research. 2014.\n\nThe cytoscape app collection. F1000Research. 2014.\n\nGao J, Zhang C, van Iersel M, et al.: F1000Research/bridgedb.cytoscape. ZENODO. 2014. Data Source" }	[ { "id": "5304", "date": "10 Jul 2014", "name": "Benno Schwikowski", "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\nBridgeDB app provides Cytoscape users and app developers with powerful functionality to address the ID mapping problem in integrative data analysis. App user and command interfaces are well-documented; the workflow is functional and recommended for anyone wishing to learn how to use the app.As a minor suggestion, it would be reassuring to new users if they were able to validate the result of the workflow example in some way.This is a very useful extension of Cytoscape functionality with broad appeal.", "responses": [] }, { "id": "5522", "date": "22 Jul 2014", "name": "Olivier Elemento", "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 Gao et. al. describes a Cytoscape app/plugin for cross-database mapping of gene, protein and metabolite identifiers in biological networks. For example, the app can map EntrezGene identifiers to UniProt identifiers for all genes or selected genes in a Cytoscape network. This is an important and useful application that among other application enables the merging of networks obtained from different sources (as shown in demonstration). The app is reasonably user-friendly and intuitive. Most importantly, it does not require any programming knowledge and is accessible via a GUI. I appreciate that it either uses local downloaded files for mapping (for heavy usage) or BridgeDb web services (for smaller networks). The demo in the paper worked flawlessly on my MacBook Air. Potential future improvements may include ortholog mapping e.g. from mouse to human to enable cross-species network analyses.", "responses": [] } ]	1	https://f1000research.com/articles/3-148
https://f1000research.com/articles/3-147/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "enhancedGraphics: a Cytoscape app for enhanced node graphics", "authors": [ "John H. Morris", "Allan Kuchinsky", "Thomas E. Ferrin", "Alexander R. Pico", "Allan Kuchinsky", "Thomas E. Ferrin", "Alexander R. Pico" ], "abstract": "enhancedGraphics (http://apps.cytoscape.org/apps/enhancedGraphics) is a Cytoscape app that implements a series of enhanced charts and graphics that may be added to Cytoscape nodes. It enables users and other app developers to create pie, line, bar, and circle plots that are driven by columns in the Cytoscape Node Table. Charts are drawn using vector graphics to allow full-resolution scaling.", "keywords": [ "Cytoscape1", "2 provides support for coloring and sizing nodes and node borders based on data values stored in the Node Table. This provides an extremely useful mapping between data values and a single visual property", "but does not solve the need for more complex visualizations. Over the years", "there have been attempts to support more complex mappings of multiple data values onto node visuals in Cytoscape. These include GOlorize3", "which maps GO terms to pie charts on nodes", "GenePro4", "which visualizes groups of nodes as pie charts", "VistaClara5", "which adds bar graphs to represent expression data", "and more recently MultiColoredNodes6. Each of these plugins and apps implemented their own graph and chart capabilities that are not accessible to other apps and in some cases not applicable outside of specific types of analyses. We felt that a better approach would be to implement an app that provided general support for graphs and charts to users as well as to the developers of other apps. The initial version of this approach was developed for Cytoscape 2.8 as nodeCharts", "which was used by clusterMaker7", "for example", "and numerous users (see Figure 3 in the Jäger", "et al. paper8). For Cytoscape 3", "we reimplemented this approach as an app to take advantage of the new architecture and custom graphics API. The mechanism supports saving and restoring charts", "as well as high-quality image file output suitable for publication. The intent is to provide a single", "consistent", "mechanism to draw charts and graphs on nodes as a general solution for diverse users and other app developers", "mitigating the need to reinvent this capability in future apps." ], "content": "Introduction\n\nCytoscape1,2 provides support for coloring and sizing nodes and node borders based on data values stored in the Node Table. This provides an extremely useful mapping between data values and a single visual property, but does not solve the need for more complex visualizations. Over the years, there have been attempts to support more complex mappings of multiple data values onto node visuals in Cytoscape. These include GOlorize3, which maps GO terms to pie charts on nodes; GenePro4, which visualizes groups of nodes as pie charts; VistaClara5, which adds bar graphs to represent expression data; and more recently MultiColoredNodes6. Each of these plugins and apps implemented their own graph and chart capabilities that are not accessible to other apps and in some cases not applicable outside of specific types of analyses. We felt that a better approach would be to implement an app that provided general support for graphs and charts to users as well as to the developers of other apps. The initial version of this approach was developed for Cytoscape 2.8 as nodeCharts, which was used by clusterMaker7, for example, and numerous users (see Figure 3 in the Jäger, et al. paper8). For Cytoscape 3, we reimplemented this approach as an app to take advantage of the new architecture and custom graphics API. The mechanism supports saving and restoring charts, as well as high-quality image file output suitable for publication. The intent is to provide a single, consistent, mechanism to draw charts and graphs on nodes as a general solution for diverse users and other app developers, mitigating the need to reinvent this capability in future apps.\n\n\nImplementation\n\nAs part of the visual property mechanism, enhancedGraphics utilizes the Cytoscape 3 custom graphics API (org.cytoscape.view.presentation.customgraphics). To use the gradients and charts provided by enhancedGraphics, an app or user would create two things: a column that contains the instructions for creating the chart, and a passthrough visual mapping that maps that column to one of the custom graphics visual properties. The format of the instruction column is type: arglist, where type is the type of gradient or chart, and arglist is a list of name=value pairs that specify the arguments to create the gradient or chart (see details and examples in the tables below). The drawing and display of the chart or graph is handled by enhancedGraphics methods that are called by the Cytoscape rendering engine.\n\nInternally, each enhancedGraphics chart type implements a CyCustomGraphicsFactory that is registered with OSGi9. Each CyCustomGraphicsFactory informs the visual mapping mechanism of the chart type (e.g. lingrad) and method to create the CyCustomGraphics object given a String, which is the instruction column value. The CyCustomGraphics object parses the String as appropriate. Each CyCustomGraphics object implements a getLayers method that generates the appropriate list of CustomGraphicLayers. The API defines three types of CustomGraphicsLayers: (1) the base interface, CustomGraphicsLayers, that provides a getPaint method to return a simple java.awt.Paint for the node; (2) ImageCustomGraphicLayer, that adds a getPaint method that returns a java.awt.TexturePaint suitable for painting an image on a node; and (3) PaintedShape which adds methods to return Shapes, Strokes, and Paints to draw arbitrary shapes. enhancedGraphics utilizes the base CustomGraphicsLayer for the two gradient types and PaintedShape for all of the charts.\n\n\nResults\n\nFigure 1 shows examples of all of the gradients and charts that are provided by enhancedGraphics. Up to nine different graphs can be combined on a single node by mapping different columns to different Custom Graphics properties and then offsetting the charts using the corresponding Custom Graphics Position properties. enhancedGraphics currently provides two different types of graphics options: gradients and charts.\n\nGradients are simple paints on nodes. In order to provide the user with control over the exact paint to use, both linear and radial gradients allow the user to specify the gradient start and end (or center point and radius) and a stop list of color and opacity values. Table 1 provides the prefixes and arguments for the two gradient types.\n\nenhancedGraphics currently provides six chart types: bar, circos, heat strip, line, pie, and stripe. Each chart type has it’s own set of arguments as shown in Table 3. In addition, there are a number of common options that are used by many of the charts. Table 2 provides the syntax and explanation for each of these common arguments.\n\nThe example charts shown in Figure 1 and provided in the Cytoscape session file Supplementary File 1 are generated from data columns. The instructions in the chart columns assume that the following columns exist: a, b, c, and d are integer columns in the default node table; Values is a list of Doubles also in the default node table, and Circle1 and Circle2 are also lists of Doubles. At this point, gradients are not dependent on any internal data. See Supplementary File 1 to see the instructions that generated Figure 1.\n\nA more relevant biological example is shown in Figure 2. This image shows a portion of the galFiltered.cys network delivered as part of the sampleData with every Cytoscape download. The bar charts show the values of the expression data included as columns gal1RGexp, gal4RGexp, and gal80Rexp. A string column was created and all rows were filled with the enhancedGraphic arguments:\n\n\n\n\nConclusions\n\nenhancedGraphics fills an important need for Cytoscape visualizations: the ability to display more complicated data relationships as graphical representations onto nodes. enhancedGraphics has been integrated into clusterMaker to show heatstrips on nodes corresponding to clusters, into upcoming apps such as cddApp, which connects to the NCBI CDD Database and uses enhancedGraphics to show pie charts of the domain coverage. We have also used enhancedGraphics to show sequence coverage histograms on nodes that represent sequence contigs. In the future, we want to improve the font handling and add support for small heatmaps painted on nodes. We also plan to add a graphical interface to help users construct these visualizations without having to write out instruction arguments. The concise syntax, however, will still be valuable to advanced users and other app developers seeking to create enhanced graphics. With enhancedGraphics, Cytoscape users and app developers can visualize multiple columns of data as graphs and charts on their network nodes.\n\n\nSoftware availability\n\nSoftware available from: http://apps.cytoscape.org/apps/enhancedGraphics\n\nLatest source code: https://github.com/RBVI/enhancedGraphics\n\nSource code as at the time of publication: https://github.com/F1000Research/enhancedGraphics/releases/tag/v1\n\nArchived source code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.1042111\n\nSoftware license:\n\nLesser GNU Public License 3.0: https://www.gnu.org/licenses/lgpl.htmlM", "appendix": "Author contributions\n\n\n\nJHM, AK, and ARP wrote the original nodeCharts plugin. TEF and ARP supervised the project and contributed to the manuscript. JHM ported the app to Cytoscape 3 and wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nARP and JHM were funded by NIGMS grant P41-GM103504. JHM and TEF were funded by NIGMS grant P41-GM103311.\n\n\nSupplementary material\n\nExample of the Instructions used to create Figure 1.\n\n\nReferences\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models ofbiomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCline MS, Smoot M, Cerami E, et al.: Integration of biological networks and gene expression data using cytoscape. Nat Protoc. 2007; 2(10): 2366–2382. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarcia O, Saveanu C, Cline M, et al.: GOlorize: a Cytoscape plug-in for network visualization with Gene Ontology-based layout and coloring. Bioinformatics. 2007; 23(3): 394–396. PubMed Abstract \| Publisher Full Text\n\nVlasblom J, Wu S, Pu S, et al.: Genepro: a Cytoscape plug-in for advanced visualization and analysis of interaction networks. Bioinformatics. 2006; 22(17): 2178–2179. PubMed Abstract \| Publisher Full Text\n\nKincaid R, Kuchinsky A, Creech M: VistaClara: an expressionbrowser plug-in for cytoscape. Bioinformatics. 2008; 24(18): 2112–2114. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWarsow G, Greber B, Falk SS, et al.: ExprEssence--revealing the essence of differential experimental data in the context of an interaction/regulation net-work. BMC Syst Biol. 2010; 4: 164. 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\nJäger S, Cimermancic P, Gulbahce N, et al.: Global landscape of HIV-human protein complexes. Nature. 2012; 481(7381): 365–370. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOSGi Alliance. OSGi service platform: release 3, March 2003. IOS Press: Ohmsha, Amsterdam; Washington, DC. 2003. Reference Source\n\nIdeker T, Ozier O, Schwikowski B, et al.: Discovering regulatory and signalling circuits in molecular interaction networks. Bioinformatics. 2002; 18(Suppl 1): S233–240. PubMed Abstract \| Publisher Full Text\n\nMorris JH, Kuchinsky A, Ferrin TE, et al.: F1000Research/enhancedGraphics. ZENODO. 2014. Data Source" }	[ { "id": "5308", "date": "17 Jul 2014", "name": "Lars Juhl Jensen", "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:The authors describe a Cytoscape App, which provides a generic interface for displaying additional data onto the nodes of a graph. I believe that this App will be highly useful to researchers working with transcriptomics and proteomics data, as these often have a need to display, for example, expression time-courses onto protein interaction networks. This is especially true if the App is used as the foundation for other more specialized Apps that make it easier for users to import and visualize specific data types in an appropriate manner.Specific comments:The Integration section is difficult to fully follow for people who are not Cytoscape developers, who cannot be assumed to know what, for example, a CyCustomGraphicsFactory is. It would in my opinion be good to revise this section to make it understandable to a broader audience.In Figure 1, I find it strange that the pie charts are not circular and that the circos charts, unlike all the other chart types, are not embedded within a frame with rounded corners. Also, the figure would benefit from a figure caption explaining the difference between Pie Chart 1 and Pie Chart 2, as well as between Circos 1 and Circos 2. The differences between these chart subtypes are not clear to me.", "responses": [] }, { "id": "6001", "date": "04 Sep 2014", "name": "Piet Molenaar", "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 app described in the article addresses a specific issue with the Cytoscape network visualization tool: drawing of multidimensional data on nodes is currently not possible by default. As such this is a very valuable addition to the Cytoscape app ecosystem. The article is well written but I do have some minor issues that can be considered for further improvement:Minor issueThe implementation details are a too technical when an intended audience includes biologists (and might even scare them away). I would suggest to transfer these OSGI specific details to a supplement or add a protocol section before this section with a sample protocol (including installation details).Minor issueThe current sample session file included in the supplemental data only contains example nodes. A session file or additional network showing actual data mapped on the gal dataset (as in the figure) would further clarify the actual usage of the app.Minor issueIn the conclusions section references to cddApp and NCBI CDD Database are lacking.", "responses": [] } ]	1	https://f1000research.com/articles/3-147
https://f1000research.com/articles/3-144/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "KEGGscape: a Cytoscape app for pathway data integration", "authors": [ "Kozo Nishida", "Keiichiro Ono", "Shigehiko Kanaya", "Koichi Takahashi", "Shigehiko Kanaya", "Koichi Takahashi" ], "abstract": "In this paper, we present KEGGscape a pathway data integration and visualization app for Cytoscape (http://apps.cytoscape.org/apps/keggscape). KEGG is a comprehensive public biological database that contains large collection of human curated pathways. KEGGscape utilizes the database to reproduce the corresponding hand-drawn pathway diagrams with as much detail as possible in Cytoscape. Further, it allows users to import pathway data sets to visualize biologist-friendly diagrams using the Cytoscape core visualization function (Visual Style) and the ability to perform pathway analysis with a variety of Cytoscape apps. From the analyzed data, users can create complex and interactive visualizations which cannot be done in the KEGG PATHWAY web application. Experimental data with Affymetrix E. coli chips are used as an example to demonstrate how users can integrate pathways, annotations, and experimental data sets to create complex visualizations that clarify biological systems using KEGGscape and other Cytoscape apps.", "keywords": [ ".bioinformatics", "cytoscape", "keggscape", "data integration", "pathway data" ], "content": "Introduction\n\nKyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg)1 is a widely used biological database of high-level biological functions. It contains pathway data sets that have comprehensive annotations and high quality human-curated, hand-drawn diagrams. Most biological pathway databases store data as machine-readable graph topologies, which leave much of the details about how the diagrams were drawn excluded from the data files. This is a problem when third-party developers want to reproduce the pathway diagrams in their applications. In contrast, the KEGG PATHWAY database stores graphics information in machine-readable KEGG Markup Language (KGML, http://www.kegg.jp/kegg/xml) format. Thus, in these pathway diagrams, biological entities, such as enzymes or compounds, are manually laid-out and the diagrams are easy to understand for biologists.\n\nThe KEGG PATHWAY database is deployed as a web application using static bitmap images for pathway diagrams, and user-provided date is integrated with KEGG Mapper (http://www.genome.jp/kegg/mapper.html). Furthermore, KEGG Atlas (http://www.genome.jp/kegg/atlas.html) provides a comprehensive network view of global metabolic pathways. Recent improvements to KEGG Atlas, such as Pathway Projector2 and iPath23, have made it possible to perform basic data integration and visualization like mapping the expression values to node graphics. However, despite these features, it is difficult to integrate external data sets and create custom visualization. Furthermore, they are limited to those on existing desktop pathway analysis applications. To ameliorate these problems, several projects for integrating a user’s own models onto the KEGG pathways have therefore been developed (CytoSEED Cytoscape app4, KEGGtranslator5).\n\nCytoscape6,7 is a de-facto standard software platform for biological network analysis and visualization. One of its advantages is its large collection of apps for a variety of biological problem domains, such as Gene Ontology term enrichment analysis (BiNGO8) and statistical network analysis (CentiScaPe9), which are also mostly open source software. Additionally, Cytoscape has a flexible network visualization function and is optimized for large-scale network analysis. There are several applications dedicated to biological pathway analysis (Vanted10, VisANT11) that support KGML by default. Although Cytoscape does not have a built-in function to load biological pathways, if this task is done with a separate app, users can take advantage of its large-scale network analysis features, variety of analysis apps, and data visualization of biologists-friendly human curated pathways.\n\nThe goal of our new Cytoscape app, KEGGscape, is to bridge the flexibility of fully-featured network analysis platforms with the high-quality pathway diagrams available in the KEGG PATHWAY web application. KEGGscape, a successor of KGMLReader (http://apps.cytoscape.org/apps/kgmlreader) for Cytoscape 2 series, is an app that imports KEGG pathway diagrams from KGML files and provides a new way to use KEGG pathway diagrams as data integration blueprints in cooperation with Cytoscape core features and an existing variety of apps. KEGGscape is completely re-designed for the new Cytoscape 3 API and supports signaling pathways in addition to metabolic pathways, including the global metabolic pathways used in KEGG Atlas (Figure 1). In this paper, we present a basic design and implementation of KEGGscape and an example workflow utilizing KGML files, and experimental data to create information-rich pathway visualizations for clarifying omics-scale data sets. KGMLReader is the first open-source Cytoscape app that reads the graphics details of KGML files, and KEGGscape was designed to use standard Cytoscape features only. These feature enable users to use KEGG pathways with other data sets easily.\n\n\nImplementation\n\nKEGGscape is a Cytoscape 3 app written in Java programming language and is designed to load pathway data files in KGML format. KGML is an XML file format designed by the KEGG project and contains the topology of pathways and visual representations of all elements in the diagram. KGML has formal specification as a DTD (Document Type Definition) file, which enables the use of unmarshaller (https://jaxb.java.net) for converting XML elements directly into Java objects. This conversion creates two types of data: pathway topology and its graphical representations. Pathway topology and its properties are converted into CyNetwork and CyTable objects, which are the standard data model in Cytoscape 3. In KGML, all graphical information, such as the color of enzymes or shape of compounds is stored under <graphics> tag. Instead of setting the graphics details of nodes and edges directly from this information, Cytoscape generates Visual Style, which is a collection of default visual properties and visual mapping function, for each pathway based on the information under this tag. KEGGscape follows a standard CyNetworkReader design guideline, which enables Cytoscape to detect KGML files automatically.\n\nFigure 2 shows an example of a pathway analysis workflow with KEGGscape. To take advantage of the flexible visualization and analysis features in Cytoscape, users need to import as much information as possible for the pathways they want to analyze. Although Cytoscape is a powerful tool for biological data integration, it is not the best platform for data preparation or cleansing. Users can instead prepare annotations and experimental data sets for the pathway using tools of their choice, such as R (Bioconductor12), Python, or Excel. Once the data files are ready, Cytoscape can read them into an on-memory session and visualize the data on the KEGG pathways. Imported data sets only use standard Cytoscape data objects, and users can then access all of the standard Cytoscape features to create custom pathway visualizations. An actual workflow will be presented in a later section.\n\nCytoscape with KEGGscape can be used as a part of larger workflows to publish integrated pathway visualizations as vector graphics, bitmap images, or JSON for web-based visualization using Cytoscape.js (http://cytoscape.github.io/cytoscape.js/).\n\nAlthough KEGGscape can read all information of the pathways saved in KGML files, some of the pathway visualizations in Cytoscape look slightly different from the original hand-drawn diagrams available on the KEGG website. The cause of this issue is missing graphics information in the KGML files. Figure 3 is a side-by-side comparison of the same pathway visualization (human MAPK signaling pathway; KEGG ID: hsa04010). The original diagram (left) contains several background visual annotations that are not visible in the visualization created by Cytoscape (right). The hand-drawn compartmental annotations are not encoded in KGML files, which means they cannot be reproduced by KEGGscape.\n\n\nResults\n\nAs an example workflow, we integrated and visualized a KEGG pathway and gene expression profile using KEGGscape and external tools. In this example, the differentially expressed genes between two groups, mutants and controls, in a global expression profile are mapped on the KEGG pathway, as too are the t-test results.\n\nTo perform this pathway analysis in Cytoscape, we used Bioconductor (http://www.bioconductor.org/) to prepare the gene expression matrix data. We normalized Affymetrix GeneChip data by the robust multi-array average (RMA) method with the Bioconductor packages ecoliLeucine13 and affy14. The leucine regulatory protein (Lrp) is a DNA binding protein and known as a leucine responsive global regulator15. The p-value for each probeset between four lrp mutant strains and four control chips was calculated by rowttest method in genefilter package16. From this calculation, we obtained a list of genes that are differentially expressed (p-value < 0.05). We sent these probeset identifiers to KEGG Mapper and picked the highest hit, which was the glycine, serine and thereonine metabolic pathway (KEGG ID: eco00260) for visualization.\n\nTo create a visualization using all the data sets, we imported the KGML file of eco00260 and the p-value matrix file prepared in the previous data preparation to Cytoscape 3, and merged the matrices with a custom Python script (Figure 4). Because Cytoscape does not support fuzzy key matching, we used our Python script to append a key column to the p-value matrix to utilize the Cytoscape table merge tool.\n\nGreen border nodes are KO (KEGG Orthology) annotated nodes. Red colored nodes include differentially expressed genes (p-value < 0.05).\n\nThe node table in Cytoscape for the imported KGML had KEGG gene annotations. The gene IDs for each enzyme node were used as keys for merging the KGML node table and p-value matrix. In this Figure 4, node colors in the original KEGG pathway were mapped to node border colors and p-values were mapped to node color gradient (red to white) to visualize the significantly expressed genes.\n\n\nConclusions\n\nIn this paper, we presented the design and implementation of KEGGscape and an example analysis workflow integrating global gene expression profiles and KEGG pathways using KEGGscape and two external tools, Bioconductor and Python. The workflow demonstrates how users can integrate omics data in an interactive pathway diagram.\n\nCurrent workflow can map arbitrary omics data onto interactive KEGG pathway diagrams, but it requires some manual editing to create informative visualizations. To minimize the manual process in the workflow, we plan to implement a collection of utility Python scripts to manipulate networks and Visual Styles via RESTful API, which will be published as a part of the Cytoscape 3.2.0 release. This set of Python scripts works to merge pathway related table metadata (omics profiles, non-KEGG pathway metadata) from external platforms like R and Cytoscape to automate common tasks in the visualization process.\n\n\nSoftware availability\n\nThe app website: http://apps.cytoscape.org/apps/keggscape\n\nLatest source code: https://github.com/idekerlab/KEGGscape\n\nSource code as at the time of publication: https://github.com/F1000Research/KEGGscape/releases/tag/V1.0\n\nArchived source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.1056017\n\nLicense: Apache License Version 2.0", "appendix": "Author contributions\n\n\n\nSK guided the GeneChip data preparation and the statistical tests, KT guided the total system design of the project. KN designed and implemented the software under KO’s supervision and performed sample analysis and visualization. KO contributed to the writing of this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by National Bioscience Database Center (NBDC) of the Japan Science and Technology Agency (JST).\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 the Google Summer of Code program and Biohackathon attendees for helpful suggestions at the early stage of KEGGscape development, and Peter Karagiannis for reading this article and for useful comments.\n\n\nReferences\n\nKanehisa M, Goto S, Sato Y, et al.: Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res. 2014; 42(Database issue): D199–D205. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKono N, Arakawa K, R Ogawa, et al.: Pathway projector: web-based zoomable pathway browser using KEGG atlas and Google Maps API. PLoS One. 2009; 4(11): e7710. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYamada T, Letunic I, Okuda S, et al.: ipath2.0: interactive pathway explorer. Nucleic Acids Res. 2011; 39(Web Server issue): W412–W415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeJongh M, Bockstege B, Frybarger P, et al.: CytoSEED: a Cytoscape plugin for viewing, manipulating and analyzing metabolic models created by the model SEED. Bioinformatics. 2012; 28(6): 891–892. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWrzodek C, Dräger A, Zell A: KEGGtranslator: visualizing and converting the KEGG PATHWAY database to various formats. Bioinformatics. 2011; 27(16): 2314–2315. 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\nSmoot ME, Ono K, Ruscheinski J, et al.: Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011; 27(3): 431–432. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaere S, Heymans K, Kuiper M: BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics. 2005; 21(16): 3448–3449. PubMed Abstract \| Publisher Full Text\n\nScardoni G, Petterlini M, Laudanna C: Analyzing biological network parameters with CentiScaPe. Bioinformatics. 2009; 25(21): 2857–2859. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRohn H, Junker A, Hartmann A, et al.: VANTED v2: a framework for systems biology applications. BMC Syst Biol. 2012; 6(1): 139. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu Z, Chang YC, Wang Y, et al.: VisANT 4.0: Integrative network platform to connect genes, drugs, diseases and therapies. Nucleic Acids Res. 2013; 41(Web Server issue): W225–W231. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGentleman RC, Carey VJ, Bates DM, et al.: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004; 5(10): R80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGautier L: ecoliLeucine: Experimental data with Affymetrix E. coli chips, 2007. R package version 1.5.0. Reference Source\n\nGautier L, Cope L, Bolstad BM, et al.: affy--analysis of Affymetrix GeneChip data at the probe level. Bioinformatics. 2004; 20(3): 307–315. PubMed Abstract \| Publisher Full Text\n\nHung SP, Baldi P, Hatfield GW: Global gene expression profiling in Escherichia coli K12. The effects of Leucine-responsive regulatory protein. J Biol Chem. 2002; 277(43): 40309–40323. PubMed Abstract \| Publisher Full Text\n\nGentleman R, Carey V, Huber W, et al.: genefilter: methods for filtering genes from high-throughput experiments. R package version 1.47.5. Reference Source\n\nNishida K, Ono K, Kanaya S, et al.: F1000Research/KEGGscape. ZENODO. 2014. Data Source" }	[ { "id": "5316", "date": "08 Jul 2014", "name": "Matthew DeJongh", "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\nKEGGscape is valuable Cytoscape app that enables flexible visualization and manipulation of KEGG pathway maps. The capability for visualizing data sets (e.g., gene expression levels) on pathways in Cytoscape is particularly appealing. However, the example presented in the paper does not contain sufficient information to enable a reader to duplicate the process. First, the authors do not cite the origin of the gene expression dataset they use in their example. Second, as far as I can tell, there is no tutorial provided with KEGGscape. At a minimum, the authors should provide instructions for opening a KGML file and loading a dataset.", "responses": [] }, { "id": "5317", "date": "22 Jul 2014", "name": "Egon Willighagen", "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 an interesting and relevant Cytoscape plugin; it is well written (there are a few typos), but I think it can be improved in places. An important reservation is that I cannot determine at this moment how common or serious the data loss in KGML reading is (see below). That is it should be clarified whether this paper is a first report, a work in progress or a finished project..Title and AbstractI find the title appropriate, but it may be promising more than what the app actually does. As far as I can tell, the app is not a general solution, but one specific for KEGG. The abstract is short on setting the state of the domain and, like the Introduction, does not refer to existing research in fields beyond their own. In this way, the abstract is less than clear what unsolved problem was resolved.IntroductionThe Introduction sufficiently explains the setting of the plugin. One thing that strikes me is the lack of references to related work. For example, it mentions \"most biological pathway databases\" without citing or even naming them, which I find suboptimal as they critique those databases and to which the authors contrast their solution. Unlike the Introduction currently seems to suggest, KGML is not the only pathway format that tracks graphical information (e.g. GPML used by PathVisio). The first paragraph currently suggests it solves a problem that others have been tackling too. (P.S. under what license is the KGML specification available?).The Introduction focuses on technical issues of the integration. It could also say a bit about identifier mapping and the role of licensing in data integration. That is, given the enormous impact of KEGG it can be wondered why this kind of integration with KEGG has not been done yet, because people have done this for other data sets for years (both open source and proprietary). Or, if it has, it should be cited.Minor comments:Check the hyphenation of KEGGtranslator (KEGG-translator versus KEG-Gtranslator).The comment that Cytoscape does not have a built-in function to load biological pathways probably refers to formats, not the pathways themselves.Since the app focuses on KGML support, some reference to apps supporting other formats (BioPax, SGML) seems appropriate.FiguresThe figures are too hard to read, both in print (where it is impossible) but also in the PDF. This should really be addressed.ImplementationThe Implementation sections can be improved: the section is short on details on the KEGGscape source code, the design of the app and the KGML reader in particular, build instructions, development model, testing (tests seem absent in the code repository).I also note issues with the KGML readers. For example, connections clear on the KEGG website are missing and misplaced when read into Cytoscape (e.g. between glycine and sarcosine, and the link between Purine metabolism, see Fig. 4) and labels are often misleadingly placed and often unconnected (a limitation of KEGG), though accurately copied from the KGML, it seems. The first data-loss clearly indicates a problem with the reader. These things must be discussed in the Limitations subsection, in my opinion (and/or returned to in the Future Plan at the end).Minor comments:on-memory -> in-memory?<graphics> tag -> <graphics> element (a tag does not really have something under it in XML, attributes at best; an element does).What is \"Visual Style\"?Under what license is the DTD available?ResultsThe Results section seems to accurately describe what they did, but not how they did this. I think something like a methods section (or an Open Notebook) that explains how the steps are performed (in more detail) would be helpful. However, in the end I was able to figure out how to import a KGML file (which is not under Import).Another example is reference to a custom Python script without description. In fact, thinking about this again, for an app that does data integration, I would actually expect the app to take care of data processing as much as possible. While I understand that general preparation starting from raw data is best done in other tools, this Python script does not seem to do more than format handling of some kind. Is that correct? If so, why is this not done by the app? Fortunately, this observation was made by the authors too, and they return to it in the Future plan subsection.Minor comments:probeset -> probe setthereonine -> threonineConclusionThe Conclusion section does not sufficiently summarize the paper: it should reiterate what the problem is that is solved and what makes KEGGscape unique. Although I do not want to imply that I think this paper is not relevant (it is)!Also the limitations should be returned to here (those already listed in the paper, and those missing), particularly in relation to the \"Future plan\". The authors’ focus should really be on getting the KGML reading performing well: data loss, and data corruption are serious issues. A decent testing toolkit as Maven supports, may be a good start here, but is currently missing from the source code repository.Minor comments:check the hyphenation of KEGGscape (KEGG-scape versus KEG-Gscape)", "responses": [] }, { "id": "5318", "date": "06 Aug 2014", "name": "Melissa Cline", "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\nKEGG is one of the foremost sources of metabolic pathway data. Cytoscape is the de facto standard pathway visualization platform. Cytoscape-based visualization of KEGG pathways has always been cumbersome and limited. KEGGscape gets over many of these limitations by streamlining the process, and improving the translation of the graphical elements of KEGG pathways. This makes KEGGscape a very useful resource for the scientific community.The manuscript is clear and informative. There are only two changes I would ask for:First, the authors should mention that KGML files are currently freely available on the KEGG website. This will be news to some readers of this paper, since KEGG has gone through various distribution and licensing models. Second, the authors should expand their discussion of merging gene expression data with KGML diagrams. The power of network visualization is in pathway data integration. If there are special challenges in integrating with KEGG pathways, these should be brought forth. The mention of custom python scripts and fuzzy ID matching suggests that this integration task can involve special challenges.", "responses": [] } ]	1	https://f1000research.com/articles/3-144
https://f1000research.com/articles/3-142/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "Development and use of the Cytoscape app GFD-Net for measuring semantic dissimilarity of gene networks", "authors": [ "Juan J. Diaz-Montana", "Norberto Diaz-Diaz", "Norberto Diaz-Diaz" ], "abstract": "Gene networks are one of the main computational models used to study the interaction between different elements during biological processes being widely used to represent gene–gene, or protein–protein interaction complexes. We present GFD-Net, a Cytoscape app for visualizing and analyzing the functional dissimilarity of gene networks.", "keywords": [ "The avalanche of information that scientists have faced during the last few years in the “-omics” fields", "has made it essential to have an appropriate computational model to run automated analysis on huge datasets1. Gene networks have arisen as a straightforward way of representing the interaction between different elements during biological processes. Gene-gene and protein-protein interaction networks have become a widely accepted way of studying how sets of proteins participate together in different biological processes2", "and multiple inference methods have been developed during the past years3–6. However", "those inferred networks must be validated in order to verify their quality and reliability." ], "content": "Introduction\n\nThe avalanche of information that scientists have faced during the last few years in the “-omics” fields, has made it essential to have an appropriate computational model to run automated analysis on huge datasets1. Gene networks have arisen as a straightforward way of representing the interaction between different elements during biological processes. Gene-gene and protein-protein interaction networks have become a widely accepted way of studying how sets of proteins participate together in different biological processes2, and multiple inference methods have been developed during the past years3–6. However, those inferred networks must be validated in order to verify their quality and reliability.\n\nGFD-Net provides a novel approach to assessing the functional dissimilarity of a gene network, i.e. the degree of dissimilarity between its genes, taking into account the relationships between them defined by the network topology. As is well known, genes may have more than one function in the organism. GFD-Net is based on an adaptation of GFD7. It uses Gene Ontology (GO)8 in order to find the most cohesive (common and specific) function of each gene based on the overall performance of the entire network. Then, it weighs each edge according to the dissimilarity between the two nodes, i.e. how close their selected functions are, and calculates a numerical value of the dissimilarity of the whole network. This value reveals the \"goodness\" or \"quality\" of the network and shows in which way the genes are closer to each other according to the information contained in GO, helping researchers to identify the overall function of the network and how each gene participates in it.\n\nCurrently, there are two main approaches for gene network validation: a direct comparison between the inferred network with gene-gene interaction repositories9 and gene annotations of biological entities10. At present there are different techniques to analyze the semantic similarity of a set of genes or gene-products11. However, to the authors’ knowledge, none of them take into account how such genes are related to each other. GFD-Net provides a new approach that also takes into account the network topology and has the advantage of constant improvement, as more specific terms are added to GO over time.\n\nGFD-Net has been integrated in Cytoscape12 as a plugin (versions 2) and as an app (versions 3). Cytoscape is a software platform for the visualization and analysis of networks, specializing in biological networks. It provides a user-friendly interface which allows users with limited software programming knowledge to use complex algorithms and computational techniques. It also has a wide range of apps13 which provide the user with the opportunity to obtain or modify a gene network using any existing app and then analyze it using GFD-Net. The large user base of Cytoscape and its apps provides the latter with a much higher visibility within the research community than they would have if they were released as stand-alone programs.\n\nIn this paper, we present the implementation of GFD-Net app for Cytoscape 3 and two simple use cases.\n\n\nImplementation\n\nGFD-Net is implemented in java and its only dependency is a JDBC driver which allows it to connect to the Gene Ontology database.\n\nFirstly, GFD-Net provides different dialogs to configure the database connection details (url, user and password), the ontology to use during the analysis, and the organism to which the network being analyzed belongs to.\n\nNext, the Cytoscape network is parsed and stored in memory using our own optimized structure for searching and quick access. The gene products associated to each gene are retrieved according to the Entrez database14, the relevant GO-terms, and the relevant section of the GO-Tree15 are loaded. Each of the proteins can be associated with, or located in one or more cellular components and be active in one or more biological processes where it can perform several molecular functions. Each annotation is represented in GO by a GO-term.\n\nGFD-Net then computes all the possible combinations of GO-terms associated to each gene in the network and tries to find the most cohesive one. Next, each edge is weighted by the dissimilarity between the selected GO-terms for the nodes at its ends, and the whole network is weighted by the average of the edge weights. Both the weights and the network dissimilarity values range from 0 to 1, where 0 and 1 represent the best and the worst values respectively.\n\nFinally, in order to facilitate the user’s interaction with the information retrieved, a result panel is displayed on the right side allowing the user to visualize all the obtained information by simply interacting with the network or the panel itself. The results are displayed in a way that allows the user to get general information about the network, or more specific information about each relationship or gene.\n\nMore details about how GFD-Net works can be found on the GFD-Net website: http://juanjoDiaz.github.com/gfdnet.\n\nOriginally, GFD-Net was a Cytoscape 2 plugin, but as soon as Cytoscape 3 was launched we ported it to an app following the Simple App approach which uses the app API to make the development similar to the old plugins. This approach requires no knowledge of the Cytoscape 3 architecture and allows a plugin to be ported with a minimal number of changes in the code but presents the same issues existing on Cytoscape 2 and its plugins. For this reason, we ported the code to a Bundle app better exploiting the benefits of the new architecture based on OSGi microservices16 and relying on Maven17 for dependency control and build instructions.\n\nGFD-Net is built following the mediating-controller MVC architecture which modularizes the code better, simplifying the maintainability of the project. By using this architecture, the app can be updated easily. For example, if the Gene Ontology database changes, or we decide to offer GFD-Net as a web service using Cytoscape.js only the data access layer or the view layer respectively will need to be modified. Figure 1 provides an overview of GFD-Net architecture.\n\nThe areas in green are directly extending or using the Cytoscape API.\n\nThe Model is completely independent of Cytoscape. It contains the application logic, the business objects and the data access layer. Since we need to traverse through a section of the GO-Tree that might be fairly large, the main challenge during the development of GFD-Net was the performance of the app. Thus, the data access layer is implemented so all the data extracted from the database is cached in memory to avoid redundant calls to the database. Furthermore, all the objects and structures used are optimized for minimal memory usage and quick searches. The retrieved data, such as genes, gene-products, GO-terms, etc., is cached in sorted sets so there are no duplicates and a specific element can be found quickly by using a binary search when needed.\n\nThe View is the layer that relies most heavily on Cytoscape’s swing application API. On the network views provided by Cytoscape the viewmodel API is used to hide or show nodes as necessary, and the model API events are used to capture the user interactions. The extensions that Cytoscape add are built using Swing and divided in two groups. The configuration dialogs are plain JDialog and provide a user-friendly interface to configure GFD-Net. The results panels are JPanels implementing the CytoPanelComponent interface in order to integrate the GFD-Net Panels in the Cytoscape UI.\n\nThe Controller gets notified of changes in the views, makes the necessary calls to the model and updates the views accordingly, completely decoupling the View from the Model. It contains actions, managers and tasks. The actions extend the AbstractCyAction class provided by the swing application API to display the menus and buttons. The managers control the different aspects of the application. There are managers to control the toolbar buttons (through the actions), the results panels, the network interactions and the core algorithm. They create the different views when necessary and are notified of user gestures on the View. Finally, the manager needs to communicate with the model to perform different operations or retrieve the content of the views. On Java Swing, everything that happens through an event (clicking a button, pressing a key, etc.) is processed by the event dispatcher thread. This means that any other event will be stuck until the current process ends and the whole UI will be blocked. Tasks extending the AbstractTask class provided by the work API of Cytoscape are run in secondary threads avoiding this issue when long running tasks are executed. Of course not all our tasks take long enough to make it necessary to use a task, so some of the calls to the model are done directly to the model. Tasks are especially important when preloading an organism (see GFD-Net website) or running the GFD-Net algorithm. Both processes can be slow (2–3 min.). GFD-Net disables all its buttons during task executions to avoid user modifications to the parameters while the program is working.\n\n\nResults\n\nGFD-Net provides an intuitive way of running a functional dissimilarity analysis on a gene network. It can be found in the Apps menu, and in order to get started, a network should already be loaded; otherwise an error will be displayed. GFD-Net adds buttons to the Cytoscape toolbar to configure the database connection, set the ontology, set the organism (preloading it or not), run an analysis and refresh the app loading the current network as selected. These buttons open the different configuration dialogs which are very user-friendly and do not require any additional details. Once all the parameters have been set, clicking on the execute button starts the analysis. When the analysis is completed, a tabbed panel showing the results is displayed on the right.\n\nIn order to show the usefulness of GFD-Net, we have analyzed two networks extracted from human pathways from Kegg18 using Graphite19; a tool found in the Bioconductor R package. Both networks can be found in the Dataset as plain text files. In both cases we configured GFD-Net the same way: online GO database (release of May 2014), Biological Process ontology and Homo Sapiens organism (without preload).\n\nFirst, we analyzed the “Cardiac muscle contraction” pathway and obtained a dissimilarity value of 0.06 (see Cardiac muscle contraction analysis results summary in the Dataset) confirming that the network has a very high functional similarity. Looking into the GO-terms associated with each gene (see Cardiac muscle contraction analysis results summary in the Dataset), we can find that the same annotation, GO:0030049 (muscle filament sliding), has been selected for all the nodes, and that many of them have annotations related to cardiac processes. It is important to note that the selected function is directly related to the pathway being evaluated proving the benefits of selecting the most cohesive set of input annotations in order to find what a networks does in the organism.\n\nThen, we analyzed the “Dorso-ventral axis formation” and obtained a dissimilarity value of 0.32 (see Dorso-ventral axis formation analysis results summary in the Dataset). At first sight, this value might not be as low as expected but the results panel in Figure 2 or in the Dorso-ventral axis formation analysis results summary in the Dataset explains the reason. The network is divided in two sub-networks (see Figure 2). The one containing SOS1, SOS2, GRB2, EGFR and KRAS is highly cohesive and all its genes have the same annotation selected, GO:0007411 (axon guidance), which is directly related with the pathway. The second one contains the nodes MAPK1, MAP2K1, MAPK3 which also have selected GO:0007411, but also ETS1 which has selected GO:0048870 (cell motility) and ETS2, ETV6 and ETV7 which have selected GO:0030154 (cell differentiation). The two later annotations show more generic functions and do not add much information about the network function, producing a higher dissimilarity.\n\nIt shows how the more specific genes are highly related while the more generic ones are not.\n\n\nConclusions\n\nWe have developed GFD-Net, a Cytoscape app that allows evaluating gene networks by finding the most common function among its genes, weighting of its edges and obtaining a value of is functional dissimilarity, as well as providing an easy way to visualize the results. As a Cytoscape app, it has the advantageous ability to interact with the broad range of existing apps. In addition, it is worth noting that GFD-Net will improve over time as more specific terms are added to gene ontology.\n\nWe have shown here, how GFD-Net provides researchers with an easy way to validate their inferred networks and find out in which way the genes in a network are related to each other. This information helps finding high functionally related subsets as well as the function of a specific gene in a given network.\n\nLooking forward, it is important to note that GFD-Net is not only restricted to being used for evaluating existing networks, but it can also be used in a gene network inference algorithm to extract more accurate models. In this line, we would expose some of the methods of GFD-Net as an API so we can have multiple apps, or multiple algorithms incorporating it. It is also in our plans to add methods to use GFD-Net directly from the Cytoscape command line. In this way we could run Cytoscape headlessly and use it as backend for a Cytoscape.js20-based website offering GFD-Net as a service.\n\n\nData and software availability\n\nF1000Research: Dataset 1. GFD-Net use cases Dataset, 10.5256/f1000research.4573.d3043723\n\nSoftware available from:\n\nApp store http://apps.cytoscape.org/apps/gfdnet\n\nApp website http://juanjoDiaz.github.com/gfdnet\n\nLatest source code https://github.com/juanjoDiaz/gfdnet\n\nSource code as at the time of publication https://github.com/F1000Research/gfdnet\n\nArchived source code as at the time of publication http://dx.doi.org/10.5281/zenodo.1062524\n\nLicense Apache License, Version 2.0", "appendix": "Author contributions\n\n\n\nJD designed and implemented GFD-Net and wrote the paper. ND conceived the idea and supervised the project. Both authors read, edited and approved the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was partially supported by the Ministry of Science and Innovation, projects TIN2011-28956-C02-1, and Pablo de Olavide University.\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\nEisenberg D, Marcotte EM, Xenarios I, et al.: Protein function in the post-genomic era. Nature. 2000; 405(6788): 823–6. PubMed Abstract \| Publisher Full Text\n\nHarrell M, Xia J, Zhao Z: Network analysis of gene fusions in human cancer. BMC Bioinformatics. 2013; 14(Suppl 17): A13. Publisher Full Text \| Free Full Text\n\nHecker M, Lambeck S, Toepfer S, et al.: Gene regulatory network inference: data integration in dynamic models-a review. Biosystems. 2009; 96(1): 86–103. PubMed Abstract \| Publisher Full Text\n\nBorelli F, de Camargo R, Martins D, et al.: Gene regulatory networks inference using a multi-GPU exhaustive search algorithm. BMC Bioinformatics. 2013; 14(Suppl 18): S5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartínez-Ballesteros M, Nepomuceno-Chamorro IA, Riquelme JC: Discovering gene association networks by multi-objective evolutionary quantitative association rules. J Computer Systems Sci. 2014; 80(1): 118–136. Publisher Full Text\n\nNepomuceno-Chamorro I, Azuaje F, Devaux Y, et al.: Prognostic transcriptional association networks: a new supervised approach based on regression trees. Bioinformatics. 2011; 27(2): 252–258. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDíaz-Díaz N, Aguilar-Ruiz JS: GO-based functional dissimilarity of gene sets. BMC Bioinformatics. 2011; 12: 360. PubMed Abstract \| Publisher Full Text \| Free Full Text\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\nWei Z, Li H: A markov random field model for network-based analysis of genomic data. Bioinformatics. 2007; 23(12): 1537–1544. PubMed Abstract \| Publisher Full Text\n\nNepomuceno-Chamorro IA, Aguilar-Ruiz JS, Riquelme JC: Inferring gene regression networks with model trees. BMC Bioinformatics. 2010; 11: 517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPesquita C, Faria D, Falcão AO, et al.: Semantic similarity in biomedical ontologies. PLoS Comput Biol. 2009; 5(7): e1000443. 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\nSaito R, Smoot ME, Ono K, et al.: A travel guide to Cytoscape plugins. Nat Methods. 2012; 9(11): 1069–1076. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaglott D, Ostell J, Pruitt KD, et al.: Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res. 2005; 33(Database issue): D54–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee SG, Hur JU, Kim YS: A graph-theoretic modeling on GO space for biological interpretation of gene clusters. Bioinformatics. 2003; 20(3): 381–388. PubMed Abstract \| Publisher Full Text\n\nOSGi Alliance. Osgi alliance \| main/osgi alliance. Retrieved: 24/5/2014. Reference Source\n\nThe Apache Software Foundation: Maven - welcome to apache maven. Retrieved: 24/5/2014. Reference Source\n\nKanehisa M, Goto S, Hattori M, et al.: From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Res. 2006; 34(Database issue): D354–D357. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSales G, Calura E, Cavalieri D, et al.: graphite - a Bioconductor package to convert pathway topology to gene network. BMC Bioinformatics. 2012; 13: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCytoscape Consirtorium. Cytoscape.js. Retrieved: 24/5/2014. Reference Source\n\nCarbon S, Ireland A, Mungall CJ, et al.: AmiGo: online access to ontology and annotation data. Bioinformatics. 2008; 25(2): 288–289. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKamburov A, Grossmann A, Herwig R, et al.: Cluster-based assessment of protein-protein interaction confidence. BMC Bioinformatics. 2012; 13: 262. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDiaz-Montana JJ, Diaz-Diaz N: GFD-Net use cases Dataset. F1000Research. 2014. Data Source\n\nDiaz-Montana JJ, Diaz-Diaz N: F1000Research/gfdnet. ZENODO. 2014. Data Source" }	[ { "id": "5760", "date": "10 Oct 2014", "name": "Cristina Rubino Escudero", "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 describes the design, implementation and use of GFD-Net, a tool to assess the functional dissimilarity of a gene network and visualize information about the function of each gene in the network.Overall, the paper is well written and provides a sound improvement on quality scoring of inferred gene networks. The abstract and keywords are appropriate and the workflow is clear. The architecture section provides useful information about how the different APIs provided by Cytoscape are use to integrate the app in Cytoscape. Finally, the use cases are well presented, easily reproducible and are a good proof-of-concept for picking most cohesive functions, proving how useful the tool can be by hinting some potential usages of this app in real biological problems.As it is mentioned in the conclusion, I think that GFD-Net full potential can be unveiled by exposing the core algorithm as an API so other apps can use it in order to extract information or as a fitness function.", "responses": [] }, { "id": "6388", "date": "03 Nov 2014", "name": "Alexander Pico", "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 the latest port and usage of GFD-Net as a Cytoscape 3 app. The calculation of GO-based functional dissimilarity (GFD) on networks provides a useful way to assess and annotate inferred networks. As part of the calculation, each pairwise interaction is weighted, providing a more granular assessment of a given network. The app takes care of mapping from gene identifiers to GO terms, the GFD calculation and the interactive display of results. The authors also share their future plans to expose an API so other apps can call on GFD-Net as a service. A welcome idea.I particularly appreciated the thorough Architecture section. Together with the open source code availability, this description will be helpful to future Cytoscape app developers interested in network model query performance, accessing GO resources and overall app design.A minor suggestion to include in your next revision of the paper: The programming language, Java, should be capitalized (first sentence in Implementation).", "responses": [] } ]	1	https://f1000research.com/articles/3-142
https://f1000research.com/articles/3-141/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "Enrichment Map – a Cytoscape app to visualize and explore OMICs pathway enrichment results", "authors": [ "Ruth Isserlin", "Daniele Merico", "Veronique Voisin", "Gary D. Bader", "Ruth Isserlin", "Daniele Merico", "Veronique Voisin" ], "abstract": "High-throughput OMICs experiments generate signals for millions of entities (i.e. genes, proteins, metabolites or any measurable biological entity) in the cell. In an effort to summarize and explore these signals, expression results are examined in the context of known pathways and processes, through enrichment analysis to generate a set of pathways and processes that is significantly enriched. Due to the high redundancy in annotation resources this often results in hundreds of sets. To facilitate the analysis of these results, we have developed the Enrichment Map app to visualize enrichments as a network. We have updated Enrichment Map to support Cytoscape 3, and have added additional features including new data formats and command line access.", "keywords": [ "With the expansion and accessibility of a wide range of experimental techniques to accurately identify and measure any known genomics feature ranging from proteins", "transcripts", "genes", "microRNAs", "copy number variations", "or DNA methylation in a high-throughput manner", "signals for thousands of entities are often generated for an individual OMICs experiment. In efforts to interpret these results in the context of perturbed cellular mechanisms", "the entities are often scored and examined for enrichment in known pathways and processes." ], "content": "Introduction\n\nWith the expansion and accessibility of a wide range of experimental techniques to accurately identify and measure any known genomics feature ranging from proteins, transcripts, genes, microRNAs, copy number variations, or DNA methylation in a high-throughput manner, signals for thousands of entities are often generated for an individual OMICs experiment. In efforts to interpret these results in the context of perturbed cellular mechanisms, the entities are often scored and examined for enrichment in known pathways and processes.\n\nPathway enrichment analysis helps to uncover general trends or themes present in the data, instead of focusing on one or a few favorite differential genes. Available tools are abundant, designed for varying data types and implemented using a range of different statistical tests: given a set of biological entities, these OMICs signals are then translated into a set of significant pathways and processes (reviewed in Khatri et al.1, Huang et al.2). Due to the high redundancy that exists between pathway databases coming from multiple functional annotations of gene products, pathway enrichment often results in a long list of potentially interesting pathways. To help analyze the set of differential pathways, we created the Enrichment Map app to display enrichment results as a network, where pathways are nodes in the network and edges represent known pathway cross-talk defined by the number of genes shared between the pair of pathways and where the network layout organizes the map into functional modules3.\n\nIn this paper, we present the recent implementation of the Enrichment Map app for Cytoscape 3 as well as new features.\n\n\nImplementation\n\nAlthough originally designed to support Gene Set Enrichment Analysis (GSEA)4 the current Enrichment Map app supports multiple enrichment results from tools such as DAVID5, BiNGO6, and GREAT7 as well as simplified generic input files which one can easily create from your own enrichment results. Tools like g:Profiler8 allow users to download results in an Enrichment Map compatible generic format.\n\nWith the ongoing effort to populate gene annotation and pathway databases, it is difficult for standalone enrichment tools to keep databases up to date. For convenience, we compile gene set files or GMT files, a format created for the GSEA software, to describe all the genes contained in a specified gene set, monthly, from a comprehensive set of annotation and Pathway databases (http://download.baderlab.org/EM_Genesets/), including standard sources, like MSigDB4. Although originally GMT files were specific to GSEA, with the expansion of R and Bioconductor it is now straightforward to load GMT files into data structures in R using packages like GSA (http://statweb.stanford.edu/~tibs/ftp/GSA.pdf) and analyze your OMICs expression data with one of the many different gene set enrichment algorithms such as geneSetTest in the Limma package9, global test10, or Camera11. Visualizing the resulting enrichments is straightforward by exporting to our generic format which minimally consists of the geneset name, description and associated enrichment p-value. Through this mechanism, no matter what the dataset of interest is, gene, protein or metabolite expression, the resulting enrichment analysis can be displayed as an enrichment map.\n\nThere are two main ways to input data into Enrichment Map, through the user interface (Figure 1) or the command tool (Table 1). The user interface is an interactive way to specify all the required files and parameters based on the analysis type chosen. The command tool allows users to automatically create maps directly from the command line, other Cytoscape apps or other programs which can include in-house enrichment tools.\n\nIllustration of Enrichment Map user interface which consists of four main parts: analysis type, file specifications, node and edge filtering. For each analysis type there is a different set of required files. For added functionality there are a set of optional files that can be included to help annotate and explore results. Tuning parameters such as p-value and q-value helps control the number of nodes while tuning the similarity coefficient helps control the number of edges.\n\nThere is an additional command optimized for GSEA inputs only.\n\nOnce files and parameters have been specified, the Enrichment Map can be created. Unlike a traditional biological network, nodes in an Enrichment Map represent a set of genes (e.g. a pathway) and their connections the set of genes that two nodes have in common (e.g. pathway cross-talk). Every Enrichment Map is associated with a set of files, parameters, and a number of datasets (currently limited to two) (Figure 2). Datasets contain gene sets, enrichments, and expression all of which is needed to interactively update the map through cutoff adjustment sliders found in the legend panel or display the genes contained in a given node or edge selection as a heatmap.\n\nEnrichment Map app was ported to Cytoscape 3 as a bundle app using Open Service Gateway initiative (OSGi) services provided through the extensive Cytoscape API (version 3.1). The look and feel of the app remains similar to the original implementation for Cytoscape 2 with user input interfaces and view panels including expression heatmap and legend being a direct port from the original source. Given the new framework, each panel implements the CytoPanelComponent and is a registered service associated with the Enrichment Map app. The main enrichment map input panel is registered only once a user opens the app. The remaining view panels are only registered once an enrichment map is created. Enrichment Map consists of one main taskFactory that given an Enrichment Map object populated with a set of input files will construct the appropriate task iterator. Depending on the files specified different parsing tasks can be added to the iterator. Additionally, multiple files of the same type can also be added to the queue with distinct instantiations of a parsing task (with different files specified on task creation). All parsed files populate fields contained in the Enrichment Map object which is then passed to and updated by each of the subsequent tasks (Figure 2).\n\nThe BuildEnrichmentMapTaskFactory is used by both the user interface and command tool to construct an enrichment map. Command tool functionality for Enrichment Map requires the given task to define its variables as tunables. Tunables are user supplied information needed by the task. User interfaces can be automatically generated for such tasks based on the set of tunable definitions. When implementing the Enrichment Map tunable task it was our intention to replace our current user interface with the one automatically generated by the task. Given the varied data required from the user as well as the interactive nature of our current user interface the generated tunable interface although functional lacked features that our users are accustomed to. For instance, to specify the analysis type or similarity cutoff our interface has two sets of radio buttons where all the options are visible and only one is selectable. In the tunable interface the same choice can only be represented as a single selection list, a drop down list the user can choose one option from. Both representations are functional but we preferred the radio button implementation therefore, we decided to keep our original interface and add the tunable task solely for the command tool functionality.\n\n\nResults\n\nTo illustrate the functionality of Enrichment Map we analyzed and visualized an expression dataset from the Gene Expression Omnibus (GEO)12 for mouse fibroblast cells. The experiment was designed to compare gene expression in fibroblast cells in the heart to those in the tail to highlight genes that are uniquely expressed in heart fibroblasts13 (GSE50531). Raw expression data was scored using the GEO2R tool available on the GEO website. These expression data were input to GSEA along with a recent compilation of mouse pathway gene sets (May 14, 2014; http://download.baderlab.org/EM_Genesets/May_14_2014/) to calculate enrichments. GSEA output files were given to the app with the cutoffs p-value < 0.005, q-value < 0.05 and overlap similarity coefficient > 0.3. The Enrichment Map generated had roughly the same number of enriched gene sets specific to heart as to tail with cardiac specific sets associated only with the heart phenotype (Figure 3, red nodes).\n\nUsing the search field you can enter any text to search all attributes of the given network. Highlighted nodes, (shown as yellow nodes with red edges just left of center) are genesets that contain the gene TBX20.\n\nOne of the main genes mentioned in the paper associated with this dataset was TBX20 as a specific cardiogenic fibroblast gene found to be important for both normal cardiac development and postinfarct repair13. In Enrichment Map it is easy to find all gene sets that contain it by entering the term TBX20 into the search box (Figure 3) (this will also highlight any gene sets that have TBX20 in the name or any other attribute). Built-in search functionality in Cytoscape 3 has improved from Cytoscape 2. All attributes associated with a given network are indexed so there is no longer the need to specify which attribute you would like to search through. Selection of individual or sets of nodes and edges creates a view of the genes contained within the selection as a heat map (Figure 4).\n\nIf GSEA results are loaded into Enrichment Map, GSEA leading edge genes, defined as the set of genes that contribute most to the enrichment, are highlighted in yellow.\n\nOften one of the main challenges after creating an Enrichment Map is going from a network in Cytoscape to publication quality figures. We format the labels so they are more readable and don’t extend across the whole screen, but as a result modules often contain overlapping labels that are difficult to read and require hours of manual formatting to create networks that can be used for figures. Using the Cytoscape 3 built-in scaling feature (Layout>Scale), the visualization of clusters and networks can be improved.\n\n\nConclusions\n\nThe Enrichment Map app allows users to translate large sets of enrichment results to a network where highly similar terms cluster together to better highlight overall trends and themes of the underlying data. The details behind the enrichment can be further investigated within the Enrichment Map app using the built-in expression viewer to see all the entities associated with a selected pathway.\n\n\nSoftware availability\n\nSoftware available from: http://apps.cytoscape.org/apps/enrichmentmap\n\nLatest source code: https://github.com/BaderLab/EnrichmentMapApp\n\nSource code as at the time of publication: https://github.com/F1000Research/EnrichmentMapApp/releases/tag/V1.0\n\nArchived source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.1054214\n\nLicense: Lesser GNU Public License 2.1: https://www.gnu.org/licenses/old-licenses/lgpl-2.1.html\n\nTutorials http://baderlab.org/Software/EnrichmentMap#Tutorials_and_Examples", "appendix": "Author contributions\n\n\n\nDM initiated and designed the project. RI wrote the manuscript and the software. RI, VV, and DM analyzed and modified existing design. GDB supervised the project.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a NRNB grant (U.S. National Institutes of Health, National Center for Research Resources grant number P41 GM103504) to Gary D. Bader.\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\nKhatri P, Sirota M, Butte AJ: Ten years of pathway analysis: current approaches and outstanding challenges. PLoS Comput Biol. 2012; 8(2): e1002375. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuang da W, Sherman BT, Lempicki RA: Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009; 37(1): 1–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMerico D, Isserlin R, Stueker O, et al.: Enrichment map: a networkbased method for gene-set enrichment visualization and interpretation. PLoS One. 2010; 5(11): e13984. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledgebased 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\nHuang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1): 44–57. PubMed Abstract \| Publisher Full Text\n\nMaere S, Heymans K, Kuiper M: BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics. 2005; 21(16): 3448–3449. PubMed Abstract \| Publisher Full Text\n\nMcLean CY, Bristor D, Hiller M, et al.: GREAT improves functional interpretation of cis-regulatory regions. Nat Biotechnol. 2010; 28(5): 495–501. PubMed Abstract \| Publisher Full Text\n\nReimand J, Arak T, Vilo J: g:Profiler--a web server for functional interpretation of gene lists (2011 update). Nucleic Acids Res. 2011; 39(Web Server issue): W307–W315. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGentleman R, Carey V, Huber W, et al.: Bioinformatics and computational biology solutions using R and Bioconductor, volume 746718470. Springer, 2005. Publisher Full Text\n\nGoeman JJ, Van De Geer SA, De Kort F, et al.: A global test for groups of genes: testing association with a clinical outcome. Bioinformatics. 2004; 20(1): 93–99. PubMed Abstract \| Publisher Full Text\n\nWu D, Smyth GK: Camera: a competitive gene set test accounting for inter-gene correlation. Nucleic Acids Res. 2012; 40(17): e133. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarrett T, Wilhite SE, Ledoux P, et al.: NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013; 41(Database issue): D991–D995. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFurtado MB, Costa MW, Pranoto EA, et al.: Cardiogenic genes expressed in cardiac fibroblasts contribute to heart development and repair. Circ Res. 2014; 114(9): 1422–1434. PubMed Abstract \| Publisher Full Text\n\nIsserlin R, Merico D, Voisin V, et al.: F1000Research/EnrichmentMapApp. ZENODO. 2014. Data Source" }	[ { "id": "5301", "date": "14 Jul 2014", "name": "Nathan Salomonis", "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 already highly used and very useful approach for making sense of highly redundant biological enrichment results that arise from the analysis of transcriptome and genomics datasets. The added additional utility of accessing the software by command line and using alternative input formats makes this a highly accessible plugin for Cytoscape, that will clearly be widely adopted by the Cytoscape community through its availability in Cytoscape 3.0.", "responses": [] }, { "id": "5298", "date": "15 Jul 2014", "name": "Florian Markowetz", "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 is an update on the Enrichment Map introduced in 2010 by the same authors. They have extended the methodology and made it available in the newest version of Cytoscape.Gene set enrichment methods of various forms are one of the most widely used first steps to gain a global picture of which pathways or other functional units are involved in some molecular phenotype. However, it is generally very hard to make sense of the results - mostly because lists of enriched gene sets can be very long and might be due to a small set of genes that appear in many of them.This is where the Enrichment Map comes in: By making the overlap between gene sets explicit it allows users to visually explore 'enrichment themes' (clusters of overlapping gene sets). Because it is simple and informative I believe this will become the standard way to present enrichment results.", "responses": [] } ]	1	https://f1000research.com/articles/3-141
https://f1000research.com/articles/2-192/v1	19 Sep 13	{ "type": "Web Tool", "title": "Cyrface: An interface from Cytoscape to R that provides a user interface to R packages", "authors": [ "Emanuel Gonçalves", "Julio Saez-Rodriguez", "Emanuel Gonçalves" ], "abstract": "There is an increasing number of software packages to analyse biological experimental data in the R environment. In particular, Bioconductor, a repository of curated R packages, is one of the most comprehensive resources for bioinformatics and biostatistics. The use of these packages is increasing, but it requires a basic understanding of the R language, as well as the syntax of the specific package used. The availability of user graphical interfaces for these packages would decrease the learning curve and broaden their application.\n\nHere, we present a Cytoscape plug-in termed Cyrface that allows Cytoscape plug-ins to connect to any function and package developed in R. Cyrface can be used to run R packages from within the Cytoscape environment making use of a graphical user interface. Moreover, it links the R packages with the capabilities of Cytoscape and its plug-ins, in particular network visualization and analysis. Cyrface’s utility has been demonstrated for two Bioconductor packages (CellNOptR and DrugVsDisease), and here we further illustrate its usage by implementing a workflow of data analysis and visualization. Download links, installation instructions and user guides can be accessed from the Cyrface homepage (http://www.ebi.ac.uk/saezrodriguez/cyrface/).", "keywords": [ "The availability of high-throughput experimental data has led to the development of multiple computational methods to analyse these data. Arguably", "one of the most used environments is the statistical programming language R1. Multiple R packages for computational biology and bioinformatics are available in various resources such as the Comprehensive R Archive Network (CRAN). Furthermore", "Bioconductor2 provides a comprehensive collection of packages to analyse biological data developed in R. These packages are subject to stringent quality control in terms of functionality and documentation. It is an open-source project hosting 671 active and curated software packages as of September 2013." ], "content": "Introduction\n\nThe availability of high-throughput experimental data has led to the development of multiple computational methods to analyse these data. Arguably, one of the most used environments is the statistical programming language R1. Multiple R packages for computational biology and bioinformatics are available in various resources such as the Comprehensive R Archive Network (CRAN). Furthermore, Bioconductor2 provides a comprehensive collection of packages to analyse biological data developed in R. These packages are subject to stringent quality control in terms of functionality and documentation. It is an open-source project hosting 671 active and curated software packages as of September 2013.\n\nFor those not familiar with computational programming, learning R and running packages can be a time consuming task and therefore the use of intuitive graphical interfaces can enhance the usability of the tool. Cytoscape3,4 is a Java open-source framework with an intuitive graphical interface devoted to the visualization and analysis of networks. It is arguably one of the most used tools in bioinformatics, and has a variety of plug-ins to solve numerous computational biology problems. Therefore, we developed Cyrface, a plug-in for Cytoscape that facilitates an interface between any R package and Cytoscape. Cyrface is designed to integrate the major strengths of R and Cytoscape environments by providing a general Java to R interface. By linking these two environments, Cyrface allows one to use Cytoscape as a user interface for R packages and Cytoscape plug-ins in order to reach the wealth of methods implemented in R.\n\nWorkflow management systems such as Taverna5 and Galaxy6–8 can call R packages from a graphical user interface (GUI)-based interface. Taverna is a standalone Java open-source tool for the general development and execution of workflows. Galaxy is an open-source web-platform to assemble workflows based on genomic experimental data analysis. Thus, Cyrface complements Taverna and Galaxy by enhancing GUIs for R within a different environment with complementary features.\n\nRCytoscape9 is another tool that exists to link R and Cytoscape. It is a Bioconductor R package that establishes a connection between R and Java in the opposite direction of Cyrface: it supports the connection from R to Java, whereas Cyrface allows a connection from Java to R. A typical use of RCytoscape is to handle experimental data from R and transfer the biological network to Cytoscape while controlling it within R. Hence, RCytoscape and Cyrface provide complementary features.\n\nThis paper is structured as follows: Firstly, we provide a description of the implementation of Cyrface. Then, to illustrate the applicability of Cyrface, we show two existing packages, CytoCopteR10 and DrugVsDisease (DvD)11, that make use of Cyrface, and we create a simplified version of the DataRail12 workflow to process and visualize experimental data using methods available in R. Finally, we discuss on-going and future developments.\n\n\nImplementation\n\nCyrface is a Java open-source framework developed to establish the connection between Cytoscape and R. Interaction between these two different environments (invoking R within Java) is not natively supported by Java. Therefore, to achieve this Cyrface uses the external libraries RCaller (https://code.google.com/p/rcaller/) and Rserve (http://www.rforge.net/Rserve/).\n\nOn the one hand, to support the communication between Java and R, RCaller uses an R package called Runiversal that converts the R objects into an XML format, thus allowing the R objects to be read by Java.\n\nOn the other hand, Rserve establishes a TCP/IP server allowing other programs from various languages to connect to an R session and access its features. Rserve is currently being used by several mature projects, among them the Taverna workflow management system5.\n\nSupport for Rserve and RCaller libraries in Cyrface is implemented by the RserveHandler and RCallerHandler Java classes, respectively. Both classes extend the abstract class RHandler that contains the signature of all the necessary methods to establish and maintain a connection with R. Figure 1 depicts the hierarchical structure of the Java classes responsible for handling the connection between Java and R. Moreover, it depicts the connection points between these two different environments.\n\nWithin the grey box the class hierarchy of the classes responsible for establishing the connection between Cytoscape and R is represented. RHandler is an abstract Java class that is extended by RserveHandler and RCallerHandler classes that add support to Rserve and RCaller libraries, respectively. The connection from Java to R can be achieved using either RserveHandler or RCallerHandler classes, or other classes that successfully extend RHandler.\n\nCyrface software architecture can be extended to support other Java libraries that facilitate the connection between Java and R. Thereby, this structure allows one to take advantage of particular strengths of different libraries and to adapt to particular requirements of the users, for instance execute R commands automatically without requiring first to manually initiate an R session.\n\nCyrface uses another Cytoscape plug-in termed CommandTool. CommandTool offers the users the ability to script basic commands in Cytoscape, such as import, display or modify networks through a simple command line. The integration allows the users to use the simple command line of CommandTool to execute R commands within Cytoscape and visualise directly the output. On Cyrface’s homepage (http://www.ebi.ac.uk/saezrodriguez/cyrface/) we provide an example using the CommandTool console to plot several characteristics of the iris data set using the ggplot13 plotting library.\n\n\nResults and discussion\n\nA typical use of Cyrface is to provide a graphical user interface to R packages within Cytoscape. Cyrface is currently being used by two Cytoscape plug-ins, CytoCopteR10 and DvD11.\n\nCytoCopteR10 provides a simple step-by-step interface allowing users without any experience in R to use the CellNOptR (www.cellnopt.org) package and handle the input and output networks in Cytoscape. CellNOptR is an open-source software package that provides methods for building predictive logic models from signalling networks using experimental measurements.\n\nDvD11, Drug vs. Disease, is an R package that provides a workflow for the comparison of drug and disease gene expression profiles. It provides dynamic access to databases, such as Array Express14, to compare drug and disease signatures to generate hypotheses of drug-repurposing.\n\nThe packages mentioned above are two examples of the usefulness of Cyrface in capturing the strengths of two environments. On one side, R provides a wealth of bioinformatics and biostatistics packages with very comprehensive resources such as Bioconductor and CRAN. On the other side, Cytoscape facilitates a user-friendly graphical interface for network visualisation and analysis, complemented with a variety of plug-ins addressing different computational biological problems. Cyrface links these two environments by providing a way to develop user-friendly interfaces for R packages by embedding them within Cytoscape.\n\nAs an illustrative example, Cyrface provides a simple version of the DataRail12 workflow using methods implemented in R. DataRail is an open-source MATLAB toolbox that handles experimental data in a tabular format and provides methods to maximize and extract information using internal or external tools. Saez-Rodriguez et al.12 also proposed an experimental data storing format termed Minimum Information for Data Analysis in Systems Biology (MIDAS).This is a tabular format based upon the minimum-information standards that specifies the layout of experimental data files. A typical use of DataRail is to import, store and process the input information from instruments using the MIDAS format, and export it to other MIDAS compliant software.\n\nThe simplified version of the DataRail workflow implemented in Cyrface is structured in several sequential steps that allows the users to import, normalise and visualise experimental data-sets stored in the MIDAS format (see Figure 2). At any stage the users are able to export and visualise the transformed data set.\n\nThe rounded rectangles represent the MIDAS files containing the experimental data at a given state. Hexagon nodes represent functions such as load or normalise. Green identifies steps that were successfully executed and grey identifies those that were not run yet.\n\nAn extension to the workflow was subsequently added to support the CellNOptR10 model training function. CellNOptR uses the experimental data and a corresponding prior-knowledge network to generate a logic model and train it to maximise the fit with the experimental measurements. Thereby, through an intuitive graphical interface, users are able to visualise a biological network, modify it and use it to assess the quality of the fit with a corresponding data set of experimental data.\n\nThe workflow supports any network format that is supported by Cytoscape, for example the SIF format. Moreover, the workflow was extended to support the Systems Biology Markup Language (SBML) Qualitative Models (Qual) format15. SBML Qual is an extension of the SBML level 3 standard and is proposed to provide a standard representation for logic and qualitative models of biological networks. The latest specification document for SBML Qual can be found on the package homepage (http://sbml.org/Documents/Specifications/SBML_Level_3/Packages/Qualitative_Models_(qual)). Support for importing models stored in SBML Qual format is achieved using the jSBML library16 and the respective SBML Qual package. Supplementary material 1 provides a step-by-step tutorial and an example on how to use the workflow.\n\n\nConclusions\n\nHere, we present Cyrface; a bioinformatics Java library that provides a general interaction between Cytoscape and R. Cyrface offers a way to combine a friendly graphical interface within the Cytoscape environment with any R package. A GUI should benefit beginners and occasional users; as well as being useful for training and illustration purposes, it extends the accessibility of the tool to those not familiar with the R command line interface.\n\nThe Cyrface homepage (http://www.ebi.ac.uk/saezrodriguez/cyrface/) contains the link to download Cyrface, and installation and user-guide instructions. A few examples demonstrating the usefulness of the tool and the different supported libraries are also shown and explained. The source-code of Cyrface is publicly available on its Sourceforge webpage (https://sourceforge.net/projects/cyrface/) and permanently available on 10.5281/zenodo.7096.\n\nFuture features for Cyrface will include the extension to the new version of Cytoscape, Cytoscape 3, and improvements to the DataRail workflow. These will include increasing its modularity and supporting other features, such as cutting and selecting specific regions of the data.\n\nHomepage: http://www.ebi.ac.uk/saezrodriguez/cyrface/.\n\nSource code: https://sourceforge.net/projects/cyrface/.\n\nPermanently available on 10.5281/zenodo.7096.\n\nLicense: GNU General Public License version 3.0 (GPLv3).", "appendix": "Author contributions\n\n\n\nJSR initiated and guided the project. EG designed the software architecture and implemented Cyrface. EG and JSR wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe acknowledge with thanks the financial support from the European Union through the project \"BioPreDyn\" (ECFP7-KBBE-2011-5 Grant number 289434).\n\n\nAcknowledgements\n\nThe authors would like to thank Martijn Van Iersel for helpful discussions and suggestions in the earlier stages of Cyrface development.\n\n\nSupplementary materials\n\nInstallation:\n\n1. Go to Cyrface homepage: http://www.ebi.ac.uk/saezrodriguez/cyrface/\n\n2. Follow the installation instructions;\n\nCyrface’s DataRail walkthrough:\n\n1. To start Cyrface’s DataRail workflow go to “Plugin -> Cyrface -> DataRail”\n\n\n\n2. The full workflow should now be visible.\n\n3. Right-click on the top MIDAS node and then select “Set MIDAS file…” to select the desired MIDAS file. After the MIDAS file is selected the node should turn green.\n\n4. Right-click on “Load MIDAS” and then select “Load MIDAS…” option to load the previously selected MIDAS file. After the file is loaded the node should turn green.\n\n5. After the MIDAS file is successfully loaded the second MIDAS node is now green showing that it’s ready to be normalized or visualized.\n\n6. Right-click on the respective MIDAS node and the selecting the “Plot MIDAS…” option will pop-up a plot of the data (the plot can be exported following “File -> Save R plot…”\n\n\n\n7. Right-click on the “Normalize” node to run the normalization function. A pop-up window will show up to allow the user to define the Normalization function arguments:\n\na. EC50Data: parameter for the scaling of the data between 0 and 1, default=0.5\n\nb. Detection: minimum detection level of the instrument, everything smaller will be treated as noise (NA), default to 0\n\nc. Saturation: saturation level of the instrument, everything over this will be treated as NA, default to Inf\n\n\n\n8. After normalizing the MIDAS file it can be ploted as previously and/or exported.\n\n\n\n-- Optional –\n\nCyrface’s DataRail Workflow is also linked to the CellNOptR R package allowing the users to optimize a selected prior knowledge network against the just normalized MIDAS file.\n\n9. Right-click on the “Optimize” node and select “Optimize…” function will pop-up a file browser to select the model file. Both Sif and SBML-qual formats are supported.\n\n\n\n10. The optimization may take awhile and it’s executed using the defaults values defined in CellNOptR\n\n\n\n11. Right-click on the “Optimized CNO List” will show how well the optimized model fit the data.\n\n\n\n12. For more details about the normalization function and the optimization method please visit CellNOptR package in Bioconductor or CellNOpt homepage\n\na. http://www.bioconductor.org/packages/2.12/bioc/html/CellNOptR.html\n\nb. http://www.cellnopt.org/\n\n\nReferences\n\nTeam RC: R: A Language and Environment for Statistical Computing.: GNU GLP. Reference Source\n\nGentleman RC, Carey VJ, Bates DM, et al.: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004; 5(10): R80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShannon P: 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\nSmoot ME, Ono K, Ruscheinski J, et al.: Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011; 27(3): 431–432. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolstencroft K, Haines R, Fellows D, et al.: The Taverna workflow suite: designing and executing workflows of Web Services on the desktop, web or in the cloud. Nucleic Acids Res. 2013; 41(Web Server issue): W557–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoecks J, Nekrutenko A, Taylor J, et al.: Galaxy: a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences. Genome Biol. 2010; 11(8): R86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlankenberg D, Kuster Von G, Coraor N, et al.: Galaxy: a web-based genome analysis tool for experimentalists. Curr Protoc Mol Biol. 2010; Chapter 19(Unit 19.10): 10.1–21. PubMed Abstract \| Publisher Full Text\n\nGiardine B, Riemer C, Hardison RC, et al.: Galaxy: a platform for interactive large-scale genome analysis. Genome Res. 2005; 15(10): 1451–1455. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShannon PT, Grimes M, Kutlu B, et al.: RCytoscape: tools for exploratory network analysis. BMC Bioinformatics. 2013; 14: 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTerfve C, Cokelaer T, Henriques D, et al.: CellNOptR: a flexible toolkit to train protein signaling networks to data using multiple logic formalisms. BMC Syst Biol. 2012; 6: 133. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPacini C, Iorio F, Gonçalves E, et al.: DvD: An R/Cytoscape pipeline for drug repurposing using public repositories of gene expression data. Bioinformatics. 2013; 29(1): 132–134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaez-Rodriguez J, Goldsipe A, Muhlich J, et al.: Flexible informatics for linking experimental data to mathematical models via DataRail. Bioinformatics. 2008; 24(6): 840–847. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer Publishing Company, Incorporated; 2009. Reference Source\n\nRustici G, Kolesnikov N, Brandizi M, et al.: Array Express update--trends in database growth and links to data analysis tools. Nucleic Acids Res. 2013; 41(Database issue): D987–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChaouiya C, Berenguier D, Keating SM, et al.: SBML Qualitative Models: a model representation format and infrastructure to foster interactions between qualitative modelling formalisms and tools. arxiv-web3.library.cornell.edu.. Reference Source\n\nDrager 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" }	[ { "id": "1866", "date": "09 Oct 2013", "name": "Nikolaus Schultz", "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, the authors describe the Cytoscape plugin Cyrface. Cyrface consists of two components: 1) a Java API, which is already being used by the Cytoscape plugins CytoCopteR and DrugvsDisease (both developed by the same group), and 2) a graphical user interface that connects R to Cytoscape. As a proof-of-concept to what kind of applications can be built on top this interface, the plugin also supports the MIDAS and SBML-Qual formats.The article is well written and the tool is useful to the community. However, we recommend the following changes to the article to make it more appealing to potential future users: As the new version of the Cytoscape (3.x) is becoming more widely used by the community, the authors should explicitly state that they are targeting version 2.8 with this framework. This will reduce the confusion for users who are not as familiar with Cytoscape. The tutorial in the supplementary materials helps to understand the general use case for this plug-in, but the lack of downloadable “sample” files for this example will make it harder for users to learn how to use the DataRail pipeline. We think it is important to provide example files that people can use for reproducing the figures in the manuscript. The Cyrface interaction layer with R looks helpful for programmers, but can the authors comment on how these classes are different from the default Java implementations of the RServe clients, e.g. http://rforge.net/RServe? This will help clarify why people should use Cyrface for their next project. The command line interface (commandTool) appears to be useful; but it seems that it is only capable of running commands in an isolated environment, with each command having its own session. If this is the case, can the authors comment on what the advantage of running R commands from the commandTool is compared to initiating a terminal window and running commands directly from an R shell? Are users able to, for example, pass node/edge attribute fields to the corresponding R commands? It looks like the current implementation does not support setting a different RServe location other than localhost. Although not necessary, if users are given the option to set a different RServe address within the plug-in, this will further lower the barrier for users who are not experienced with R to use Cyrface, where they can use a pre-installed Rserve hosted on a different machine.", "responses": [] }, { "id": "2157", "date": "12 Dec 2013", "name": "Ghislain Bidaut", "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 well written, however, many items require clarification:My main criticism is the initial need to install several packages (RCaller, Rserve, Runiversal, Java, etc...), plus Cytoscape and the CommandTool plugin. Since I did not know if this would break my personal configuration or if I would be able to uninstall it, I was not able to performe the whole installation myself. It would be great if the authors could provide a virtual machine image with all the software preloaded so that one can try it out of the box without installation. I know by experience that cityscape plugins tend to work only with a single version of Cytoscape, so a clear list of all the required versions in the paper itself would be very useful. The number of packages we are dealing with is very confusing. I am also afraid that with such a large number of dependencies, the program may break after any update. In the Implementation section, the authors mention the \"iris dataset\". It would be useful to define what this is. It is also mentioned in the documentation but it is still unclear what the authors are referring to. The documentation shows an example where the ggplot2 package is used to plot \"petal\". Could the authors please define what this is? Figure 2: It would be useful to define more accurately what type of data we are looking at here (e.g. gene expression?). In the discussion, several R packages are mentioned (DvD, Cytocopter), what are their links with the present software other than the fact that they run under R? Is it really possible to interact with them from Cytoscape? If so, the proper documentation or a tutorial should be provided. It would also be useful if an actual example using DvD along interaction data, could be shown. I do not understand the example with DataRail. The usefulness of the example given is not clear to me since it seems that no interaction data is given (for me this is the main purpose of Cytoscape). A useful example for most users might be:- A network of protein-protein interaction (PPI) data in Cytoscape.- Some expression data in R (an exprs object for instance).An example of a question might be \"How to superimpose the expression data and generate a proper network attribute from it?\"", "responses": [] }, { "id": "1863", "date": "20 Jan 2014", "name": "Paul T. Shannon", "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\nCyrface is a welcome addition to the Cytoscape ecosystem; nicely complementary to RCytoscape. My only reservation is one which applies to my own work (the aforementioned RCytoscape), indeed as well as all parts of the Cytoscape ecosystem.My reservation has two parts:First: network biology is in its infancy and as such experimental data are woefully incomplete. Molecular interactions are stochastic, contingent and often very short-lived yet it is exactly these molecular interactions that we need to understand in order to predict and control cellular activity in health and disease.Second: it is (and may remain) unusual to find researchers, much less clinicians, who are adept at both programming and biomedicine. These two disciplines seem to select for, and then reinforce, different styles of thinking. Therefore progress in this field (call it network biology, or systems biology, or integrative biomedicine) requires hybrid teams: some who are very strong in biological and/or clinical sensibilities and some who are strong in computation and data analysis. Such a hybrid team, at its best, stays together long enough for mutual understanding and communication to emerge, as in the \"trading language\" which emerged in the world of particle physics in and around linear accelerators in the 60's (see Peter Galison's \"Image and Logic\").I worry about the following scenario for Cyrface: a capable programmer hooks up the latest and greatest Bioconductor package to Cytoscape, exposes as best they can the parameterizations offered by that package, and turns the tool over to their collaborating biologist. Experimental data is loaded and analyses or simulations run. Puzzles and inconclusive results will inevitably emerge, requiring detailed knowledge of both the strengths and weaknesses of the Bioc package. With good luck, perseverance and good data, this small working team may in time settle on a satisfactory Cyrface tool which can be reused without the constant intervention of the programmer. This will last until new data is acquired, upsetting the equilibrium, and the hybrid style of work and the back-and-forth between biologist and programmer, begins anew.I say that I worry about this scenario. It may be exactly the intended use of Cyrface; the problem it is intended to solve. But this essentially sociological characteristic (requirement?) of Cyrface is not described in the paper. I think that those of us who create bioinformatics software tend to avoid being explicit about this - and I think that this (the social & collaborative requirements of bioinformatic research) deserves a lot more attention. If indeed network biology, as I claim, is in its infancy, then it may be helpful if the ecosystem of Cytoscape-related tools are considered from this perspective. I suspect that the conclusions we might all (mostly) agree upon are:User-friendly exploration of data-rich networks in a web browser (as with cytoscape.js) will become increasingly popular. That user-friendliness often competes with analytical nuance and close scrutiny - biologically and clinically useful results become less likely. Cyrface's connection of Cytoscape to R is a great step in the right direction, marrying as it does user-friendliness with some new analytical power in a way that is nicely complementary to Cytoscape java plugins and Cytoscape access to web services.Thus, Cyrface is a good step in the right direction. The next steps, it seems to me, include:Providing easy connections to R (python, C++) analyses for cytoscape.js A standard mechanism whereby scripts (R, python, Ruby, Perl) upon execution, can start up a Cytoscape or cytoscape.js session, customize it with networks, functions, buttons and menus, and with both public and laboratory data. As a generalization of Cyrface, this mechanism would encourage the rapid expansion of Cytoscape capabilities.These possible next steps carry on in the spirit of Cyrface, RCytoscape, and Cytoscape3 apps, and will promote the creation of, and sharing of, custom network analyses, shared tools, and lead to fruitful collaborations across the hybrid community of biologists, physicians and programmers.", "responses": [] } ]	1	https://f1000research.com/articles/2-192
https://f1000research.com/articles/2-288/v1	30 Dec 13	{ "type": "Research Article", "title": "JSim, an open-source modeling system for data analysis", "authors": [ "Erik Butterworth", "Bartholomew E. Jardine", "Gary M. Raymond", "Maxwell L. Neal", "James B. Bassingthwaighte", "Erik Butterworth", "Bartholomew E. Jardine", "Gary M. Raymond", "Maxwell L. Neal" ], "abstract": "JSim is a simulation system for developing models, designing experiments, and evaluating hypotheses on physiological and pharmacological systems through the testing of model solutions against data. It is designed for interactive, iterative manipulation of the model code, handling of multiple data sets and parameter sets, and for making comparisons among different models running simultaneously or separately. Interactive use is supported by a large collection of graphical user interfaces for model writing and compilation diagnostics, defining input functions, model runs, selection of algorithms solving ordinary and partial differential equations, run-time multidimensional graphics, parameter optimization (8 methods), sensitivity analysis, and Monte Carlo simulation for defining confidence ranges. JSim uses Mathematical Modeling Language (MML) a declarative syntax specifying algebraic and differential equations. Imperative constructs written in other languages (MATLAB, FORTRAN, C++, etc.) are accessed through procedure calls. MML syntax is simple, basically defining the parameters and variables, then writing the equations in a straightforward, easily read and understood mathematical form. This makes JSim good for teaching modeling as well as for model analysis for research.\n\nFor high throughput applications, JSim can be run as a batch job. JSim can automatically translate models from the repositories for Systems Biology Markup Language (SBML) and CellML models. Stochastic modeling is supported. MML supports assigning physical units to constants and variables and automates checking dimensional balance as the first step in verification testing. Automatic unit scaling follows, e.g. seconds to minutes, if needed. The JSim Project File sets a standard for reproducible modeling analysis: it includes in one file everything for analyzing a set of experiments: the data, the models, the data fitting, and evaluation of parameter confidence ranges. JSim is open source; it and about 400 human readable open source physiological/biophysical models are available at http://www.physiome.org/jsim/.", "keywords": [ "experimental data analysis", "partial differential and ordinary differential equations", "differential algebraic equations (DAE)", "JSim", "modeling", "optimization", "continuous systems", "stochastic systems simulation", "sensitivity", "time series", "reproducible research", "SBML", "CellML." ], "content": "Introduction\n\nThe modeling of biological processes starts with defining the hypothesis to be tested in an experiment. To make scientific progress, Platt (Platt, 1964) emphasized defining at least two distinct hypotheses and then designing an experiment with the power to clearly distinguish between these hypotheses. By so doing, at least one of the hypotheses must then be rejected: the rejection marks a stepping-stone in science. If a hypothesis is not rejected then it remains as a potential working hypothesis, the target of further experimentation that eventually will lead to its rejection or improvement.\n\nThe virtue of the mathematically-defined hypothesis is that it is clear and precise, and therefore susceptible to contradiction. Arguably, one should use mathematical “in numero experimentation” to define the critical laboratory experiment. Given that the experiment tests whether or not the working hypothesis is compatible with experimental data, then failure to fit leads to rejection. Revision of the conjecture follows: science is advanced.\n\nThe hypothesis testing cycle is an iterative procedure: design hypothesis (and alternative hypotheses) → execute experiment → evaluate goodness of fit of model to data → either reject the hypothesis and restart, or, alternatively, → accept the model as the current working hypothesis and assess the parameters for the specific situation. The working model serves as the current belief until deeper thinking leads to an alternative hypothesis and one restarts the cycle. This philosophical and procedural point of view, more or less guaranteed to make efficient progress in the field, creates definable results step by step, and gives investigators a sense of satisfactory success.\n\nAs in physics, models are posed in order to gain deeper understanding. Cause-and-effect models of biological systems are usually deterministic; they are fundamentally different from observationally-based probabilistic associations. The desire is to represent sequences of operations within a dynamic system leading to, and explaining, the observed data (Coatrieux & Bassingthwaighte, 2006; Bassingthwaighte et al., 2006a). Standard statistical methods are not central to deciding whether or not to reject the hypothesis, but are indeed very helpful in assessing goodness of fit, estimating confidence ranges and co-variances among parameters, and in guiding the investigator in identifying errors or in finding ways to simplify the model.\n\nOver the years we have developed sets of tools to serve these processes. In this article we describe the features of a simulation analysis system, JSim; it is the product of evolutionary improvements in the hypothesis testing cycle. The central goals are to facilitate attempts to fit models to data, and to support the efficient development of computational models that describe and explain the behavior of biological systems (Bassingthwaighte & Goresky, 1984; Bassingthwaighte et al., 2005; Beard et al., 2005).\n\nOur perspective is embedded in JSim: it is an open-source simulation analysis platform, freely downloadable, running on Linux, Macintosh, and Windows, providing tools for the steps in the modeling analysis of data. There is a naturally occurring sequence of steps to take when one starts with an unanalyzed data set and has the goal of modeling the cause and effect relationships. We have found it useful to follow a simplified summary: The THIRTEEN STEPS:\n\n\nThe THIRTEEN STEPS in the modeling process\n\nThese are proposed as a guide. The ordering is not rigid, but it is wise to cover all of the steps in one’s mind when starting and again when finishing up a study. Using the steps in the order listed here almost always works well.\n\n(1) When starting with existing experimental data, plot and display the data so that one can rapidly review and compare multiple data sets. This also prepares for comparing with later model results.\n\n(2) Develop the model, the mathematical formulation of the hypothesis. One may start with one or more existing models or modules of a similar nature (retrieved from a model repository or archival format) and modify it. Construct illustrations of model structure to aid the conceptual approach.\n\n(3) Verify unitary balance in the model equations, an easy first check for model self-consistency.\n\n(4) Select appropriate methods for solving model equations (e.g. differential equation solvers).\n\n(5) Display model solutions graphically and in text listings. Inspect.\n\n(6) Verify the mathematical accuracy of solutions. Check that results are not dependent on temporal or spatial step sizes, that mass or charge is appropriately conserved, and that limiting cases match analytical solutions.\n\n(7) Explore model behavior over wide ranges of parameter values in state-space. (We think of “state space” as being the N-dimensional space enclosing the ranges of values of all of the parameters within which the model is correct numerically and sensible scientifically.)\n\n(8) Perform sensitivity analyses, examining the fractional change in model solutions with fractional change in each parameter.\n\n(9) Adjust parameters to fit model solution to data, manually or using an optimizer. Start from different places in parameter space and vary the optimization method to test solution uniqueness.\n\n(10) Assess goodness of model fit to data. Plot residual differences to expose systematic biases.\n\n(11) Examine parameter correlations to identify highly correlated parameters and reduce the number of free parameters in optimizations. Reoptimize.\n\n(12) Evaluate parameter confidence ranges. The sensitivities at the “best fit”, expressed as the local curvature of the optimization cost function give a practical estimate. This can be refined using a Monte Carlo evaluation of parameter likelihoods as probability density functions.\n\n(13) Preserve the source code, multiple data sets, multiple analyses and parameter sets, the settings (for initial and boundary conditions, parameter scans, displays, solver choices, optimizers, Monte Carlo, etc.), the graphs of results, the investigator’s notes and descriptions of procedures, plots, etc., all in a single, reproducible, exportable package. Share this package openly with collaborators, reviewers, and the public, a moral and perhaps ethical requirement when the support comes from public funds.\n\nWhat one wants primarily from modeling analysis is insight into mechanisms. JSim is efficient for model development and testing. The fitting of experimental data by model solutions does not provide proof that the model is correct. It says merely that the model can serve as a descriptor under limited range of circumstances, namely those examined in the experimentation. Validity is never provable. Likewise, causation may be identified, but deeper levels may exist to be revealed later.\n\nWhat does the model predict? Every model, with a little ingenuity, can be queried. What would be the responses to different inputs? How would the system respond if a component were missing or damaged? Predictions then form the basis for the design of the next experimental test. Correct predictions, failing to invalidate the model, do strengthen the confidence in the model but only to the degree commensurate with the comprehensiveness of the particular prediction.\n\nJSim is the latest in a series of modeling/data analysis programs dating back to SimCon (Knopp et al., 1970) (named for Simulation Control). SimCon provided a text and graphics interface to models written in Fortran. Between 1967 and 1993, the basic methods of data analysis (e.g. function generators, loops, sensitivity, optimization) were developed and refined within the SimCon framework. In 1993, SimCon was replaced by XSim (King et al., 1995), which implemented the same functionality under X-Windows on several Unix-like operating systems (SunOS, IRIX, Linux, AIX). XSim also added custom graphic model interfaces, on-demand expression graphing, worlds-within-worlds graphics (Harris et al., 1994), remote (client-server) computation and limited multi-processing. JSim development efforts began in 1999 and augmented the functionality developed in SimCon and XSim by adding simplified model specification (using the MML modeling language), facilities for data analysis and for distribution of results and of models (using project files), popular desktop and laptop support (Windows, Macintosh & Linux) and fully integrated multiprocessing for shared memory systems (Raymond et al., 2003).\n\nJSim is quite general, and while designed for evaluating models against experimental data, it also serves pure model development quite well. It is built around a “project file” (.proj), that may hold many data sets, several different models and the results of multiple types of analyses testing models against the data and against each other. JSim’s handling of ODEs (ordinary differential equations) suits it for traditional compartmental modeling and SBML (Hucka et al., 2003), CellML (Cuellar et al., 2003), and pharmacokinetic (PK) models in general. Solving PDEs (partial differential equations) hugely expands the range of processes that can be modeled in physiology and clinical medicine (Goresky, 1963; Bassingthwaighte, 1974), biophysics, and PKPD modeling (Roberts & Rowland, 1986). JSim handles spatial diffusion (Barta et al., 2000; Safford & Bassingthwaighte, 1977) and convection-diffusion problems. From soon after its release in 1999, JSim provided automated unit consistency checking in all equations and also automated unit conversion (such as minutes to seconds) in calculations (Chizeck et al., 2009). This pair of features automates the first stage of verification of the model’s mathematical implementation by making sure that every equation has unitary balance. Modeling taking account of the anatomical quantitative constraints is now recognized as critical and is facilitated by the automated unit checking (Vinnakota & Bassingthwaighte, 2004). The second phase of compilation parses the details of the equations and sequences them for efficient computation. For an example, a cardiovascular-respiratory system model (Neal & Bassingthwaighte, 2007), ran under JSim exactly 300 times faster than a Matlab-Simulink version of the identical model (Howard Chizeck/Stephen Hawley: personal communication).\n\nMML (Mathematical Modeling Language) is the declarative modeling language developed for JSim and used for composing models. Its archival version is XMML, in the XML style of SBML and CellML. In MML, one writes mathematical equations directly into the code, and the MML compiler handles converting the set of equations into a sequence of computations. Since the equation representation is closely related to the conceptual formulation of the model, MML models are easily understood, and pieces of the model are readily interpretable as particular processes. The fact that one can write several models into a single MML program allows one to compare competing hypotheses (models). Having a standard layout for graphs and for ASCII text output of model solutions is convenient. For special purposes, as for a model to be used in clinical practice or teaching, an alternative graphical user interface specifically designed for the model can be readily substituted for the default layout. If a particular model absolutely requires procedural code, this can be developed in C, or Fortran or Java, and invoked as part of the model computation.\n\nJSim is a general purpose simulation and data analysis software system. It handles a wide range of mathematical problems including algebraic equations, ordinary differential equations, and parabolic, hyperbolic and elliptic partial differential equations. It contains 8 ODE and 3 PDE solvers implementing a variety of algorithms which allow the flexibility to strike a balance between accuracy and computational speed. It performs time series analyses including forward and backwards Fourier transforms. MML can handle multi-dimensional PDEs but the solvers currently implemented support only two dimensions (typically time and one spatial dimension). For two spatial dimensions the problem needs to be formulated into either ODE nodes or PDEs in one spatial dimension linked by ODEs in the other spatial dimension. JSim does not support complex numbers or matrix notation and associated matrix operators; in JSim all matrices must be written explicitly as a set of equations.\n\nJSim can be used in any discipline where mathematical equations are used for modeling and analyzing data. JSim was originally developed to model and analyze physiological phenomena and many of the built-in tools were developed to handle physiological problems. But all of the JSim tools can be applied to any other scientific discipline. JSim excels at analyzing time course and spatial domain data in complex systems (Beard & Bassingthwaighte, 2000; Beard et al., 2005; Bassingthwaighte et al., 2006b; Suenson et al., 1974; Safford & Bassingthwaighte, 1977). Examples include modeling pharmacokinetic/dynamic (PK/PD), radiological (CT, PET, MRI) and multiple indicator dilution (MID) data.\n\nJSim uses the Mathematical Modeling Language (MML) to describe models. When JSim imports other model formats (e.g. SBML, CellML, Antimony), it translates them to MML. MML is a concise, ASCII text language for defining parameters and variables and for writing the equations describing a model. MML is a declarative language (as opposed to procedural or imperative languages such as MATLAB, Java, Python, and FORTRAN), meaning that, in MML, equations represent mathematical equality, rather than providing a directive to calculate the left-hand side variable from the expression on the right. In MML, it makes no difference if terms in an equation appear on the left or right hand side. Such equations are a direct representation of the mathematical ideas in a model rather than a procedural formulation. This improves readability and allows for more extensive consistency checks than procedural formulations. The MML compiler checks to ensure that all variables are completely, but not overly, specified – a check unavailable in procedural languages. The compiler sequences the calculations based on the dependencies of the variables to be computed, thus eliminating order-of-operations errors that are possible in procedural languages. MML variables are (optionally) labeled with physical units, enabling the compiler to reject equations with unitary imbalances; this also allows the automated insertion of appropriate unit conversion factors when needed (Chizeck et al., 2009) (e.g. mmHg to kPa). This relieves the modeler of the burden of adding unit conversion factors (another potential source of error) and aids readability, since equations need not be cluttered with conversion factors. MML’s design supports the model development and unit balance aspects of modeling steps 2 and 3 above. An example of MML code is shown below as Box 1, which codes a “progress curve”, the concentration-time curves for hypoxanthine to xanthine to uric acid catalyzed by the enzyme xanthine oxidase through the two oxidation steps. MML code for partial differential equations is given in Box 2.\n\n// Model Name: MM2irrev (From reference JBBass13, data of Escribano (Escribano et al., 1988))\n\n/* Brief Description: The “MM2irrev” program codes a sequential pair of irreversible Michaelis-Menten enzymatic reactions, Hx → Xa → Ua, wherein the one enzyme, xanthine oxidase, serves both steps. Hx and Xa compete for its active site. /\n\nimport nsrunit; unit conversion on;\n\nmath MM2irrev {\n\nrealDomain t sec; t.min=0; t.max=5000.0; t.delta=1.00; // t is independent variable\n\n// PARAMETERS: (denoted param(t) if time-variable) (all changeable at run-time)\n\nreal Vhmax = 1.84 uM/s; // Vmax for enzymatic conversion of Hx -> Xa\n\nreal Kmh = 3.67 uM; // Km for assumed instant binding of Hx to enzyme\n\nreal Vxmax = 1.96 uM/s; // Vmax for Xa -> Ua\n\nreal Kmx = 5.94 uM; // Km for assumed instant binding of Xa to enzyme\n\nreal Hzero = 46.3 uM, Xzero = 0 uM, Uzero = 0 uM; // initial conditions\n\n// VARIABLES (specified as functions of time by (t) appended in defining the name)\n\nreal H(t) uM; // concentration of Hx (HypoXanthine)\n\nreal X(t) uM; // concentration of Xa (Xanthine)\n\nreal U(t) uM; // concentration of Ua (Uric acid)\n\n// INITIAL CONDITIONS (t.min can differ from t = 0 sec.)\n\nwhen (t=t.min){ H= Hzero; X = Xzero; U = Uzero;}\n\n// SYSTEM OF EQUATIONS (3 ODEs) (Derivative dH/dt written as H:t)\n\nH:t = - (VhmaxH/Kmh) / (1 + H/Kmh + X/Kmx); // Hx→Xa\n\nX:t = ((VhmaxH/Kmh) - (VxmaxX/Kmx)) / (1 + H/Kmh + X/Kmx); // Xa→\n\nU:t = (VxmaxX/Kmx) / (1 + H/Kmh + X/Kmx); // →Ua\n\n} // PROGRAM END\n\nMML is designed without reference to the numerical algorithms that will be used for simulation. Rather, the user selects the numerical methods in the JSim run time user interface. At present JSim provides 8 algorithms for solving ODEs (Table 1) and 3 for PDEs (Table 2). Numerical methods for stochastic simulation are variants on the Gillespie algorithm (Gillespie, 1977). JSim’s solvers support modeling steps 4 to 6 above.\n\nTo solve differential equations one needs initial conditions, and JSim’s parser (precompiler) demands these, as in Box 1. Partial differential equations require also boundary conditions, as seen in the code for a two-region convection-diffusion-permeation-reaction model (Box 2).\n\nMany physiological systems or components (e.g. one for the uptake of a metabolite) can be considered as operators. The operator takes an input function (e.g. inflowing solute concentration) and produces an output function (e.g. outflowing solute and metabolite concentrations). Model behavior can be tested by using various input waveforms (e.g. as in Box 2 “extern real Cin(t)”) described by JSim “function generators”. These might be time series signals of diverse form (pulses, pulse combinations, sines, shaped sawtooth), probability density functions (Gaussian, exponential, Poisson, lognormal, gamma variate, random walk, etc.), or come directly from experimental data. When the system is linear (output area equals input) and stationary (response same at another time), then the output is the convolution of the operator’s transfer function (the response to an infinitely short pulse input) with the input function. Users select input functions at run time for testing numerical algorithms for correctness (verification testing), for model exploration (behavioral analysis) or for analyzing data as for steps 6 and 7 in our “13-Step” process.\n\nWe will use a simple convection-diffusion reaction model (Bassingthwaighte, 1974; Bassingthwaighte & Goresky, 1984) to illustrate some facilities for visualizing model solutions and the effect of varying parameter values on them. The system is diagrammed in Figure 1 and the code is provided in Box 2.\n\nFluid flows with velocity FcapL/Vcap along the capillary from the entrance at x = 0 to the exit at x = L, and exchanges across the capillary wall into a stagnant extravascular region with conductance PS, the permeability-surface area product. The input is a bolus of solute, Cin(t), entering the capillary with the flow, Fcap. Axial gradients along the capillary are diminished by diffusion, Dp and Disf. Tissue consumption occurs at rate GisfCisf. This is a simplified version of models used for indicator dilution studies and PET clinical studies (Beard & Bassingthwaighte, 2000; Bassingthwaighte et al., 1989; Bassingthwaighte et al., 1992; Bassingthwaighte et al., 2006b).\n\n// MODEL NUMBER: 0190 at www.physiome.org (Bassingthwaighte, 1974)\n\n// MODEL NAME: BTEX20simple\n\n// SHORT DESCRIPTION: Simple Model of an axially distributed two-region\n\n// capillary Blood-Tissue EXchange unit with consumption in interstitium\n\nimport nsrunit; unit conversion on;\n\nmath btex20simple {\n\n// INDEPENDENT VARIABLES:\n\nrealDomain t sec ; t.min = 0; t.max = 30; t.delta = 0.1;\n\nrealDomain x cm; real L= 0.1 cm, Ngrid = 31; x.min = 0; x.max = L; x.ct = Ngrid;\n\n// Parameters and Keys to Names:\n\nreal Fcap = 1 ml/(gmin), // Capillary (cap) plasma flow\n\nVcap = 0.05 ml/g, // Capillary Volume\n\nVisf = 0.15 ml/g, // Interstitial Fluid (isf) Volume\n\nPS = 1 ml/(gmin), // Permeability-surface area product: cap <--> isf\n\nGisf = 0 ml/(gmin), // consumption rate in isf region (Gulosity)\n\nDcap = 1.0e-5 cm^2/sec, // cap axial diffusion coefficient\n\nDisf = 1.0e-6 cm^2/sec; // isf axial diffusion coefficient\n\n// Inflow Concentration, Input Function:\n\nextern real Cin(t) mM;\n\n// Concentration Variables:\n\nreal Ccap(t,x) mM, // capillary concentration at position x\n\nCisf(t,x) mM, // isf concentration at position x\n\nCout(t) mM; // Outflow Concentration from capillary at x=L\n\n// Boundary Conditions: (Note total flux BC for inflowing region.)\n\nwhen (x=x.min) { (-FcapL/Vcap)(Ccap-Cin)+DcapCcap:x = 0; Cisf:x = 0; }\n\nwhen (x=x.max) { Ccap:x = 0; Cisf:x = 0; Cout = Ccap; } // reflecting boundary\n\n// Initial Conditions:\n\nwhen (t=t.min) { Ccap = 0; Cisf = 0; } // sets initial concentrations to zero\n\n// Partial Differential Equations: Ccap:t is dCcap/dt in JSim’s MML (ODE or PDE)\n\nCcap:t = -FcapLCcap:x/Vcap + DcapCcap:x:x + PS(Cisf-Ccap)/Vcap; // dCcap/dt\n\nCisf:t = -GisfCisf/Visf + DisfCisf:x:x + PS(Ccap-Cisf)/Visf; // dCisf/dt\n\n} //program end\n\nJSim provides several mechanisms for visualization, providing insight about model dynamics. The most basic are plot pages, each of which may contain line, scatter, contour and colormap plots. One may plot experimental data and model solutions (from one or more models), scaled automatically or manually, linear or logarithmic, plotted as they are being computed or displayed or edited later. Multiple plot page configurations are stored in each project, enabling reproducible analysis (e.g. all the data and graphs for a particular journal article). JSim plot pages support modeling steps 1, 5, 6, 7, 8 and 10 above (display of experimental data and model solutions, verify solution accuracy, explore model behavior, display of sensitivity curves and assessments of goodness of fit).\n\nModel loops are a feature for behavioral analysis that plot data from a family of model runs using a user-chosen sequence of parameter values. For example, Figure 2, “looping over”, i.e. making a sequence of changes in a parameter value for the membrane permeability in a tracer uptake model yields a family of plots showing how outflow tracer concentration curves would vary with varying permeability. The curves, of course, depend upon the settings for the other parameters of the model, so the looping sequence should be initiated under widely divergent conditions in order to understand the “conditions” (the regions of state space) where the chosen parameter may have little influence or maximum influence. JSim’s loops facility support modeling steps 6 and 7 above (verify solution accuracy, explore model state space). A convenient feature of the LOOPS function is that the user can stop the solution, automatically starting the next one, whenever desired, speeding up the review of solutions. This is especially important in large models with long computation times.\n\nWith the permeability surface area product (PS) = 0 (taller solid curve) the outflow concentration-time curve, Cout, represents the response function via the vascular space alone. The mean transit time for this is Vcap/Fcap. With finite PS there is extraction of solute during transcapillary passage, shown by the successive diminutions of the heights of the initial peaks as PS increases. At low PSs the form of the outflow starts as a reduced version of the curve with PS = 0; in this state the flux into the tissue is purely \"barrier limited\". When PS is 4 or greater ml/(gmin), the sixth curve, the initial peak is no longer discernable; at yet higher PSs a second peak arises, and at PSs above 128 ml/(gmin) increasing the PS further has no effect on the shape of the outflow curve; in this state the exchange flux is purely \"flow-limited\", where changing the flow shifts Cout, but changing PS does not.\n\nNested plots (Figure 3) are JSim’s version of worlds-within-worlds graphics (Harris et al., 1994). Each nested plot is a 2-dimensional array of plots, each of which represents the form of a set of model solutions with a pair of distinct parameter value. Nested plots enable simultaneous visualization of the effect of up to six independently varying parameters. JSim nested plots support modeling steps 6–8 above (verify accuracy, explore model state space, sensitivity analysis).\n\nBehavior of the two-region model when varying capillary permeability, PS, and tissue consumption, Gisf. Each panel is a contour plot with the position between the capillary entrance at x=0 to the exit at x=0.1 cm on the abscissa, and time, t, on the ordinate. At each time step (ordinate) the horizontal line from 0 to 20 is colored (using color profile “rainbow” in this case) in accord with the concentration at each point in x. Convection moves the entering solute along the tube from left to right to larger x on this graph. With successive times the colored horizontal lines construct a shaped profile above the x-t plane; contour lines with units in mM are superimposed. The columns from left to right show contours with PS increasing by factors of 5 (see labels at top of column) from PS = 0.3 to 37.5 ml/(gmin). The consumption Gisf increases from 0.2 in the bottom row by factors of 5 to the top row with Gisf = 25 ml/(gmin); see labels on right ordinate. With low PS, leftmost column, very little of the solute escapes into the tissue, so the injected bolus remains relatively compact even while undergoing some diffusional spread (Dcap = Disf = 10-4 cm2/sec), and the influence of the consumption is negligible since so little enters the ISF. With increasing PS more solute enters the ISF where it is consumed. With high PS and high Gisf, the right uppermost panel, the solute is all consumed before it can reach the capillary exit at the right edge of the panel. [This plot is set up under “Project”, “Add”, “New Nested Plot” using LOOPS, inner and outer, to set the values for the parameters, and on the NestedPlot, then clicking on “XY plot” to choose “contour”. Instructions are under Running JSim – Data Analysis – Nested plots: www/physiome.org/jsim/docs/User.html].\n\nBy “sensitivity analysis” we mean the examination of the influences of individual parameters on the model responses under a wide variety of conditions. The sensitivity function, S(t) is the change in magnitude, dQ, of variable Q, to a small change in a parameter value, dP. It may be expressed in a normalized form, (dQ/Q)/(dP/P), or unnormalized form, dQ/dP. As an example consider the same model as was explored in Figure 3. Figure 4 shows the sensitivities of the outflow concentration of a solute to a change in interstitial fluid volume (Visf) or capillary wall conductance (PS) following an injection of that solute at the capillary entrance. The upper panel shows the outflow concentration without parameter perturbation. The middle panel plots the unnormalized sensitivity functions, and the bottom plot shows the normalized sensitivity functions (with the early part of the curves removed when Cout is negligible). Increasing PS will lower the height of Cout for the first 10 seconds with the greatest reduction at the peak of Cout at ~8 seconds (due to greater flux of metabolite into the ISF); after 10 seconds, the height of Cout will be increased (back flux of metabolite from ISF). Increasing Visf has the effect of lowering Cout for the first 24 seconds, then raising it after 24 seconds. JSim’s sensitivity analysis supports modeling step 7 above.\n\nUpper panel: Model solution for outflow from capillary. Parameters were as in Box 2, the default parameters. Middle panel: Sensitivity function, df/dp, the change in Cout with a 1% increase in the capillary wall conductance (PS), black curve or the interstitial volume (Visf). Lower panel: Normalized sensitivity function, (df/f)/(dp/p), the fractional change in Cout divided by the fractional change in each parameter, again for a 1% change in the parameter value.\n\nManual parameter adjustment to fit the model to experimental data is encouraged as a means of gaining insight into model behavior. Automated parameter optimization is usually much faster; eight methods are provided (See Table 3); we recommend testing several in order to test speed and reliability with respect to the particular types of data and model. Given that some parameters are known or highly constrained, one may obtain the best model fit to the data for a particular subset of model parameters, and one may also, for some but not all of the optimizers, constrain the range for each parameter value, applying scientific judgment. Optimization helps in finding systematic misfits to the data (and the possible rejection of the hypothesis), and in estimating parameter values.\n\nThe optimizer works to minimize an objective function, usually a weighted sum of squares of the differences between the model solution and the experimental data at each observation time or spatial position. This may require freeing up most parameters for optimization to make sure that an assumed constraint isn’t creating a biased solution. JSim provides a graph of residuals (the differences between model and data); sign tests and other statistical appraisals of the residuals as a function of time help to distinguish systematic from random deviations. JSim’s optimization facilities support modeling steps 9–12 above (fitting solutions, assessing goodness of fit, examining parameter correlations, evaluating confidence limits).\n\nModel fitting to the data is never unique but is guided by the weighting of the observed data points and the noise in the data. Parameter estimates are not exact, but merely estimated, and even possibly biased by the user’s choice of the weights on individual data points. How to obtain a “best fit” of model function to data is always, in a sense, a personal choice. Guidelines include weighting inversely to the likely standard deviation of each data point, or unweighting outliers. Viewing the graph of residuals (the differences between data and model) is most helpful in identifying systematic misfits.\n\nIgnoring how one got to the point of “best fit”, one desires an evaluation of the parameter values. If the optimized parameters do generate outputs that closely match the experimental data, the question becomes what confidence can be placed on these estimates. One simple method is to optimize using several different numerical method, i.e. different optimization algorithms and different weighting schemes, to see how much the “best fit” parameter estimates change. Other methods of estimating parameter confidence limits include using the Jacobian and using Monte Carlo methods.\n\nUsing the Jacobian: The Jacobian matrix is the matrix of the sensitivity functions for all the parameters open to optimization, as calculated at the location of the minimized objective function, the “best fit”. This matrix, which JSim calculates after each optimization provides the basis for determining correlations among parameters, and the confidence limits (standard deviations and expected ranges based on Gaussian assumptions). The calculation assumes symmetry and linearity, and so makes only local calculations, and gives no guarantee that the “best fit” is a global best fit. While getting to the “best fit” point in parameter space is data-dependent, this confidence range estimation procedure is not at all, for it is estimated solely from the shapes of the local sensitivity functions. Thus it behooves one to get the differing estimates obtained from different optimizers, different numbers of parameters searched, and even to move the parameter “best fit” values a little away from the optimizer’s choice and recalculate the confidence ranges.\n\nUsing a Monte Carlo method: A more robust, but more demanding, confidence limit calculation uses Monte Carlo methods. The procedure is to 1) Select a noise profile for each experimental data point, ideally based on what you believe the real noise is, e.g. 5% proportional Gaussian random noise. 2) Generate a perturbation for each experimental data point by drawing randomly from the selected noise profile. 3) Re-optimize the model against the new set of perturbed data points to obtain another estimate for each parameter. 4) Repeat steps 2 and 3 many times (e.g. 1000). From these results, one obtains a histogram of estimates for each optimized parameter, and robust confidence limits can be drawn directly from these histograms without assuming symmetry and linearity as in the Jacobian method. JSim displays these histograms to show the distributions of parameter estimates in full detail, and 2-parameter scatter plots to show covariance. (JSim’s confidence limit calculations support modeling step 12 above.)\n\nJSim’s model “browser” provides a visual representation of model variables as “nodes” and their dependencies or connectivity with each other as connecting lines or “edges”. See Figure 5. The graphs can be selected to include model parameters, or selected classes of variables, e.g. pressures, strains, concentrations. This capability is based on work by Yngve (Yngve et al., 2007). JSim’s model browser supports modeling step 2 above (development of the model).\n\nFor Hx→Xa→Ua, the oxidation of hypoxanthine to xanthine to uric acid, catalyzed by xanthine oxidase. The connectivity is shown for a dual solution version of the code for fitting two different sets of experimental data simultaneously with a common group of parameters so as to obtain a minimally biased set of parameter confidence ranges.\n\n\nImplementation\n\nJSim is implemented in the Java computer language (Gosling & McGilton, 2003). The major factors affecting this choice are Java's platform independent GUI (allowing Windows, Macintosh and Linux versions to be developed in a single code base), object-oriented features and garbage collection (simplifying complex coding), advanced utilities (associative arrays, dynamic linking, remote procedure calls), strong type checking (allowing many common coding errors to be caught at compile time) and robust exception mechanism (simplifying coding and enabling a virtually crash-proof GUI). Native code (C and Fortran) is used in certain restricted circumstances using the Java Native Interface (JNI) (Liang, 1999) to reduce computational overhead (transcendental functions, 2D array access) and the availability of legacy code libraries (ODE, PDE and optimization numerical methods).\n\nThe MML language is parsed using JLex scanner generator and the CUP parser generator (Appel, 1998). These tools, similar to the classic Unix lex and yacc (Aho et al., 1988), were among the few parser generation tools available for Java when JSim was first developed. Using a formal parser generator allows MML to be concise, intuitive, consistent and extensible. MML's declarative structure is an intuitive expression of a model's underlying mathematics (simplifying the modeler’s learning) and allows the overall structure of the model to be examined for mathematical correctness (detecting overspecification or underspecification) in a way that is not possible with a procedural specification. Units and unit checking (Chizeck et al., 2009) were added to MML soon after its initial design to further improve model conciseness and assure unit balance in the equations as a first step in verifying that the mathematics is rendered correctly by the numerics.\n\nMML is compiled into Java model computational code for run-time execution. This results in faster model execution (in comparison to table-driven computations) and allows more flexible model computational structure (multiple time sweeps, indexed loops). JSim models run asynchronously to the GUI in contrast to most simulators which alternate computational and graphical update steps. This approach dramatically improves performance and user response, especially when remote computation is used. JSim's remote computation is implemented using Java Remote Method Invocation (RMI) (Harold, 1997), providing reliable access to networked computational servers. This approach also isolates the JNI methods (above) in the computational engine, allowing the JSim GUI to run as a pure Java browser applet. JSim multiprocessing is implemented using Java threads (Oaks & Wong, 2004) providing excellent performance and seamless integration with the Java memory management and exception mechanisms (providing application stability). MML code is stored as XMML for distribution, and has automated translators into XMML, SBML, CellML, and with limitations into Matlab (Smith et al., 2013).\n\n\nReproducibility\n\nThe issue of reproducibility, or should we say the all-too-frequent failures of attempts to reproduce published results, are beginning to be recognized as a critical problem in advancing the biological sciences. It is easy to understand biological studies, with inherently great variability in materials and analysis procedures, should be less exact than those in the physical sciences, but it is not so forgivable that reproducing mathematical models of biological systems is a major problem. The two major repositories of published biological models, Biomodels (http://www.ebi.ac.uk/biomodels-main/) using SBML (www.sbml.org) and CellML (models.cellml.org), together have about 1000 curated models: there were errors in the publications requiring corrections in all but 5 of these, before the models could be demonstrated to run appropriately. These models all used algebraic, ODEs, or differential-algebraic equations and so must be regarded as relatively simple computationally compared to finite-element models or spatially dependent models. That only 0.5% of the not very complex models were reproducible is truly alarming, and demonstrates the lack of dedication to making scientific advances useful to others. Some open access journals, such as F1000Research, are aiming to improve this sad state, by requiring open source code to be deposited, hopefully along with the data that provide tests of the model hypotheses. A Special Section in Science (Stone & Jasny, 2013) is devoted to the issues of open access, addressing open access, peer review, the changing publishing scenario, and encouraging broader methods of communication. F1000s founder, Victor Tracz, is featured as the “Seer of Science Publishing”, prodding us to do better.\n\nJSim project files store a set of codes for models, illustrative figures or diagrams, parameter sets, multiple data sets, the settings for looping, sensitivities, behavioral analysis, and optimizations, plot page configurations, and for project notes. Many models in the Physiome Repository (most of which are JSim-based) have experimental data in the project files for validation testing. Project files support the reproduction of a set of simulations and analyses for their sharing across JSim’s supported platforms (Windows, Mac OS, Linux). Project files support the modeling steps 1 and 13 above (from importation of data, to preservation and distribution of analyses). The MML, XMML and all the data and analyses are preserved in an ASCII format; thus the files tend to be small. The models described above take < 100 kB; large models with several hundred ODEs take up < 500 kB even with large time series of physiological data. These files are all human readable, and ready to run when opened in JSim. They contain everything used by the program: the notes, the source code, and the control parameters for all the steps in the analysis. They are editable in any word processor, but one avoids doing that since it is easier to enter code and notes under JSim and not risk disturbing the format in the XMML file that JSim reads.\n\nThere are many models on the Physiome Repository (www.physiome.org) with multiple data sets, model fits to data, and optimization results. Examples are that of Kuikka et al. on glucose uptake by myocardium (Kuikka et al., 1986), [models 163 and 173], xanthine oxidase reactions (Bassingthwaighte & Chinn, 2013), [model 324], and lung endothelial serotonin uptake (Jardine & Bassingthwaighte, 2013), [model 198]. All the JSim project files are stored in a Concurrent Versions System (CVS) archive so that the latest versions, as well as older versions, are always available. The models themselves are copyrighted but researchers are given the freedom to download, modify, and to construct new models from them so long as original authorship is acknowledged.\n\nThe archived JSim models at www.physiome.org can be run over the web, with complete freedom to vary the parameters, modify the code, compile and run, import one’s own data for analysis, and save a modified and augmented file to one’s own computer for further use. (Models based on MATLAB or FORTRAN, a small fraction of the repository, cannot be run over the web but can be downloaded).\n\n\nSummary\n\nJSim is a tool for hypothesis exploration and for analyzing data. Many of the steps in data analysis are built into JSim. It’s declarative modeling language, automatic unit balance checking, and built-in tools for solving ODEs, PDEs, and implicit equations greatly facilitate generating mathematically and physiologically consistent models. The built-in optimizers and associated statistical data reporting, along with behavior tools, such as parameter looping and sensitivity analysis, allow one to verify and explore model behavior in the context of experimental data and simulated data from previous models. With the ability to save these model ‘explorations’ as parameter sets within the JSim project file anyone can easily create a modeling and data analysis package that is easy to reproduce and distribute to others.\n\nAs a research tool, JSim has been developed and refined to accelerate the processes of modeling and data analysis. Adherence to quality standards augments efficiency (Smith et al., 2007). The time savings don’t simply reduce the time necessary to get to a result, they also end up improving the quality of the science in two ways. First, when it only takes a few seconds to tweak a model, re-run it, and view the results, researchers are more likely to explore many “what if” scenarios and develop a deeper understanding of model behavior, and in turn, a deeper understanding of the system being modeled. Second, researchers are more likely to do better verification checks and higher-level analyses if they are easy to do. When a few mouse clicks are all it takes to change solvers, time step sizes, optimization parameters, or even perform a complex Monte Carlo analysis to assess parameter correlations and confidence intervals, researchers are more likely to actually do those critical numerical checks and to take the model analysis beyond simply reporting a single parameter value.\n\nIn addition to its use as a research tool, JSim is also very useful as a teaching tool. JSim has been used in classes for high school, undergraduate, and graduate students, as well as many workshops for faculty members. The fact that JSim is open source, quick to download and install, as well as executable over the web, means that it is easily available to students. The simplicity of JSim’s model specification language, where users can focus on writing and working with the mathematical equations themselves rather than controlling program flow, means that students with no programming experience can rapidly begin to understand, create, and modify JSim models. Furthermore, JSim’s interactive plotting interface and the easy access it provides to sophisticated analysis tools such as sensitivity and Monte Carlo analysis allow students to perform analyses which would ordinarily be too difficult and time consuming for them to do on their own.\n\n\nFuture developments\n\nJSim has provided support for modular modeling from its inception (Bassingthwaighte, 2000) using both mathematical and biological approaches, but now, with the developing recognition that models are more consistently understandable and more amenable to modular construction when they are annotated using identified ontology systems, libraries of modules present great opportunity for efficient construction of complex model systems. A module can be thought of as a self-contained model of an element of the larger system model and represents a specific physical, chemical or phenomenological process. A model might use multiple instances of the same module, for example, differently parameterized Michaelis-Menten type enzymatic reactions used for different reactants. One can build large models from a variety of modules representing physical or chemical processes such as the flux via a cell membrane transporter or ion channel or an enzymatic reaction, or a transcription regulatory pathway (Beard et al., 2005) incorporating knowledge of their connectivities. Allowing the modeler to draw pre-existing modules from a repository or extract them from previously developed models and enables the modeler to create new models quickly for hypothesis testing, a key to Physiome development (Bassingthwaighte et al., 2009). Below are two approaches to implementing modular modeling within JSim.\n\nModular Program Constructor (MPC): MPC focuses on using easily understood directives to extract generically coded JSim MML equations from files, changing the names of the generic variables to ontologically informative names and assembling the resulting code into new equations (Raymond & Bassingthwaighte, 2011). For example, MPC can take MML code representing a single tissue exchange region (26 lines), and generate a whole organ heterogeneous model for convection, diffusion, and reaction with 20 regions (1698 lines). See http://www.physiome.org/jsim/models/webmodel/NSR/MPC/. MPC currently runs outside of JSim but is planned for incorporation into a future JSim release.\n\nModular construction with SemSim: Precise semantic identification of variables and parameters is a prerequisite to merging of preconstructed submodels or modules into integrated systems or multiscale models. A future release of JSim will incorporate the tools for annotating models and their computational elements against biomedical ontologies and knowledge bases (Rubin et al., 2006). These annotations will make it easier for users to search the Physiome Model repository and to identify the biological phenomena modeled. Formatted according to the semantic simulation (SemSim) framework (Gennari et al., 2011), these annotations will also make it possible for tools to decompose and merge models in a more automated fashion, and allow the modeler to work at a biological, rather than computational level of abstraction (Beard et al., 2012). For example, selection of an ion pump, such as the NaKATPase, would bring up a set of modules from which the modeler would choose the version suited to the particular context, and then the code for the integrated model would be automatically generated from the annotated modules in the library.\n\nInformation for download and installation, running JSim, and writing JSim MML models can be found at http://www.physiome.org/jsim/. Software is also permanently available from: 10.5281/zenodo.7635.", "appendix": "Author contributions\n\n\n\nAll authors contributed to the design and organization of the paper and its writing and editing. EB was the developer of the JSim engine and GUI. GR implemented numerical methods for numerical solvers and optimizers. BJ worked on the website and model coding, formatting and installation. MN developed large systems models and the methods of ontology annotation. JB was the overall systems designer and composed the integrated manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe development of JSim has been supported by NIH grants HL9719 (PI: JBB), RR1243 (JBB), EB1273 (JBB), HL073598 (PI: R. Corley), EB8407(JBB), and GM094503 (PI: D. Beard) and NSF grant 0506477(JBB).\n\n\nAcknowledgements\n\nThe authors would like to thank M. Bindschadler and H. Sauro for helpful discussions and feedback regarding the manuscript. Tom J. Knopp, Dennis U. Anderson, and Richard B. King contributed to the simulation methods used in the preceding simulation systems, SIMCON and XSIM, that were incorporated into JSim.\n\n\nReferences\n\nAho A, Sethi R, Ullman J: Compilers: Principles, Techniques and Tools. Addison-Wesley. 1988. Reference Source\n\nAppel AW: Modern Compiler Implementation in Java. Cambridge University Press. 1998. Reference Source\n\nBarta E, Sideman S, Bassingthwaighte JB: Facilitated diffusion and membrane permeation of fatty acid in albumin solutions. Ann Biomed Eng. 2000; 28(3): 331–345. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB: A concurrent flow model for extraction during transcapillary passage. Circ Res. 1974; 35(3): 483–503. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Goresky CA: Modeling in the analysis of solute and water exchange in the microvasculature. In: Handbook of Physiology. Sect. 2, The Cardiovascular System. Vol IV The Microcirculation, edited by Renkin EM and Michel CC. Bethesda, MD: Am Physiol Soc. 1984; pp 549–626.\n\nBassingthwaighte JB, Chan IS, Goldstein AA, et al.: GGOPT: an unconstrained non-linear optimizer. Comput Methods Programs Biomed. 1988; 26(3): 275–281. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Wang CY, Chan IS: Blood-tissue exchange via transport and transformation by capillary endothelial cells. Circ Res. 1989; 65(4): 997–1020. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Chan IS, Wang CY: Computationally efficient algorithms for convection-permeation-diffusion models for blood-tissue exchange. Ann Biomed Eng. 1992; 20(6): 687–725. PubMed Abstract\n\nBassingthwaighte JB: Strategies for the Physiome Project. Ann Biomed Eng. 2000; 28(8): 1043–1058. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Chizeck HJ, Atlas LE, et al.: Multiscale modeling of cardiac cellular Energetics. In: The Communicative Cardiac Cell. edited by Sideman S, Beyar R and Landesberg A. Ann New York Acad Sci. 2005; 1047: 395–424. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Chizeck HJ, Atlas LE: Strategies and Tactics in Multiscale Modeling of Cell-to-Organ Systems. Proc IEEE Inst Electr Electron Eng. 2006; 94(4): 819–830. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB, Raymond GR, Ploger JD, et al.: GENTEX, a general multiscale model for in vivo tissue exchanges and intraorgan metabolism. Philos Trans A Math Phys Eng Sci. 2006; 364(1843): 1423–1442. PubMed Abstract \| Publisher Full Text\n\nBassingthwaighte J, Noble D, Hunter P: The Cardiac Physiome: perspectives for the future. Exp Physiol. 2009; 94(5): 597–605. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassingthwaighte JB: Modeling biological systems for reproducibility and sharing. TOPETJ (The Open Pacing, Electrophysiology, and Therapy Journal). 2010; 3: 66–74.\n\nBassingthwaighte JB, Chinn TM: Reexamining Michaelis-Menten enzyme kinetics for xanthine oxidase. Adv Physiol Educ. 2013; 37(1): 37–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeard DA, Bassingthwaighte JB: The fractal nature of myocardial blood flow emerges from a whole-organ model of arterial network. J Vasc Res. 2000; 37(4): 282–296. PubMed Abstract \| Publisher Full Text\n\nBeard DA, Bassingthwaighte JB, Greene A: Computational modeling of physiological systems. Physiol Genomics. 2005; 23(1): 1–3. PubMed Abstract \| Publisher Full Text\n\nBeard DA, Neal ML, Tabesh-Saleki N, et al.: Multiscale modeling and data integration in the virtual physiological rat project. Ann Biomed Eng. 2012; 40(11): 2365–2378. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlom JG, Zegeling RUU: Algorithm 731: A moving-grid interface for systems of one-dimensional time-dependent partial differential equations. ACM Transactions on Mathematical Software (TOMS). 1994; 2(2): 194–214. Publisher Full Text\n\nChan IS, Goldstein AA, Bassingthwaighte JB: SENSOP: a derivative-free solver for non-linear least squares with sensitivity scaling. Ann. Biomed. Eng. 1993; 21(6): 621–631. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChizeck HJ, Butterworth E, Bassingthwaighte JB: Error detection and unit conversion. Automated unit balancing in modeling interface systems. IEEE Eng Med Biol. 2009; 28(3): 50–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoatrieux JL, Bassingthwaighte JB: Special Issue on the Physiome and Beyond. Proc IEEE. 2006; 94: 671–677. Reference Source\n\nCohen SD, Hindmarsh AC: CVODE, a stiff/nonstiff solver in C 1995, UCRL-JC-121014, (Cohen et al. ver. 2.4 Jul 2002). Reference Source\n\nCuellar AA, Lloyd CM, Nielsen PF, et al.: An Overview of CellML 1.1, a Biological Model Description Language. SIMULATION. 2003; 79(12): 740–747. Publisher Full Text\n\nDantzig GB, Orden A, Wolfe P: The generalized simplex method for minimizing a linear form under linear inequality restraints. Pacific J Math. 1955; 5(2): 183–195. Reference Source\n\nDennis JE, Gay DM, Welsch RE: NL2SOL: An adaptive nonlinear least-squares algorithm. ACM Trans Math Softw. 1981; 7(3): 348–368. Publisher Full Text\n\nDennis JE, Schnabel RB: Numerical methods for unconstrained optimization and nonlinear equation. N. Y.: Prentice-Hall, 1983; 378. Reference Source\n\nEscribano J, Garcia-Canovas F, Garcia-Carmona F: A kinetic study of hypoxanthine oxidation by milk xanthine oxidase. Biochem J. 1988; 254(3): 829–833. PubMed Abstract \| Free Full Text\n\nEuler L: Institutionum calculi integralis. 2007; 2. : 1768–1770. Reference Source\n\nFehlberg E: Low-order classical Runge-Kutta formulas with stepsize control and their application to some heat transfer problems. NASA Technical Report-315. 1969. Reference Source\n\nGennari JH, Neal ML, Galdzicki M, et al.: Multiple ontologies in action: Composite annotations for biosimulation models. J Biomed Inform. 2011; 44(1): 146–154. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGillespie DT: Exact stochastic simulation of coupled chemical reactions. J Phys Chem. 1977; 81(25): 2340–2361. Publisher Full Text\n\nGoresky CA: A linear method for determining liver sinusoidal and extravascular volumes. Am J Physiol. 1963; 204: 626–640. PubMed Abstract\n\nGosling J, McGilton H: The Java language environment: A white paper, 1995. Sun Microsystems. 2003; 85. Reference Source\n\nHairer E, Norsett SP, Wanner G: Solving Ordinary Differential Equations. Nonstiff Problems. 2nd edition. Springer Series in Comput. Math., 1993; 8. : 528. Reference Source\n\nHairer E, Wanner G: Solving Ordinary Differential Equations. Stiff and Differential-Algebraic Problems. 2nd edition. Springer Series in Comput. Math., 1996; 14. : 614. Reference Source\n\nHarold ER: Java Network Programming. O'Reilly. 1977. Reference Source\n\nHarris PA, Bosan S, Harris TR, et al.: Parameter identification in coronary pressure flow models: a graphical approach. Ann Biomed Eng. 1994; 22(6): 622–637. PubMed Abstract\n\nHolland JH: Adaptation in natural and artificial Systems: an introductory analysis with applications to biology, control, and artificial intelligence. MIT Press. 1992; 183. Reference Source\n\nHucka M, Finney A, Sauro HM, et al.: SBML Forum. The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models. Bioinformatics. 2003; 19(4): 524–31. PubMed Abstract \| Publisher Full Text\n\nJardine B, Bassingthwaighte JB: Modeling serotonin uptake in the lung shows endothelial transporters dominate over cleft permeation. Am J Physiol Lung Cell Mol Physiol. 2013; 305(1): L42–L55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKing RB, Butterworth EA, Weissman LJ, et al.: A graphical user interface for computer simulation. FASEB J. 1995; 9: A14. (XSim).\n\nKirkpatrick S, Gelatt CD Jr, Vecchi MP: Optimization by simulated annealing. Science. 1983; 220(4598): 671–680. PubMed Abstract \| Publisher Full Text\n\nKnopp TJ, Anderson DU, Bassingthwaighte JB: SIMCON--Simulation control to optimize man-machine interaction. Simulation. 1970; 14(2): 81–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKolda TG, Lewis RM, Torczon V: Optimization by direct search: New perspectives on some classical and modern methods. Siam Review. 2003; 45(3): 385–482. Publisher Full Text\n\nKuikka J, Levin M, Bassingthwaighte JB: Multiple tracer dilution estimates of D- and 2-deoxy-D-glucose uptake by the heart. Am J Physiol. 1986; 250(1 Pt 2): H29–H42. PubMed Abstract \| Free Full Text\n\nLeVeque RJ: Finite Difference Methods for Ordinary and Partial Differential Equations. Philadelphia: Siam. 2007. Publisher Full Text\n\nLiang S: The Java Native Interface: Programmer's Guide and Specification. Addison-Wesley. 1999; 303. Reference Source\n\nMacCormack RW: The Effect of viscosity in hypervelocity impact cratering. AIAA Paper. 1969; 40(5): 69–354. Publisher Full Text\n\nMerson RH: An operational method for the study of integration processes. Proc. Symp. Data Processing , Weapons Res. Establ. Salisbury , Salisbury. 1957; pp. 110–125.\n\nNeal ML, Bassingthwaighte JB: Subject-specific model estimation of cardiac output and blood volume during hemorrhage. Cardiovasc Eng. 2007; 7(3): 97–120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelder JA, Mead R: A simplex method for function minimization. Computer Journal. 1965; 7(4): 308–313. Publisher Full Text\n\nOaks S, Wong H: Java Threads (Third Edition). O'Reilly. 2004. Reference Source\n\nPlatt JR: Strong inference: Certain systematic methods of scientific thinking may produce much more rapid progress than others. Science. 1964; 146(3642): 347–353. PubMed Abstract \| Publisher Full Text\n\nPoulain CA, Finlayson BA, Bassingthwaighte JB: Efficient numerical methods for nonlinear-facilitated transport and exchange in a blood-tissue exchange unit. Ann Biomed Eng. 1997; 25(3): 547–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRaymond GM, Butterworth E, Bassingthwaighte JB: JSIM: Free software package for teaching physiological modeling and research. Exper Biol. 2003; 280(5): 102.\n\nRaymond GM, Bassingthwaighte JB: Automated modular model construction using JSim. Experimental Biology. 2011; 863(9).\n\nRoberts MS, Rowland M: A dispersion model of hepatic elimination: 1. Formulation of the model and bolus considerations. J Pharmacokinet Biopharm. 1986; 14(3): 227–260. PubMed Abstract\n\nRubin DR, Grossman D, Neal ML, et al.: Ontology-based representation of simulation models of physiology. AMIA Annu Symp Proc. 2006; 664–668. PubMed Abstract \| Free Full Text\n\nSafford RE, Bassingthwaighte JB: Calcium diffusion in transient and steady states in muscle. Biophys J. 1977; 20(1): 113–136. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith NA, Crampin EJ, Niederer SA, et al.: Computational biology of the cardiac myocyte: proposed standards for the physiome. J Exper Biol. 2007; 210(Pt 9): 1576–1583. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith L, Butterworth E, Bassingthwaighte JB, et al.: SBML and CellML translation in Antimony and JSim. Bioinformatics. 2013. (In press). PubMed Abstract \| Publisher Full Text\n\nStone R, Jasny B: Scientific Discourse: Buckling at the Seams. Science. 2013; 342(6154): 57–82. (Eight articles covering new views on improving scientific communication and accelerating research through openness and clarity.). Publisher Full Text\n\nSuenson M, Richmond DR, Bassingthwaighte JB: Diffusion of sucrose, sodium, and water in ventricular myocardium. Am J Physiol. 1974; 227(5): 1116–1123. PubMed Abstract \| Free Full Text\n\nVinnakota KC, Bassingthwaighte JB: Myocardial density and composition: a basis for calculating intracellular metabolite concentrations. Am J Physiol Heart Circ Physiol. 2004; 286(5): H1742–H1749. PubMed Abstract \| Publisher Full Text\n\nYngve G, Brinkley JF, Cook D, et al.: A model browser for biosimulation. AMIA Annu Symp Proc. 2007; 2007: 836–840. PubMed Abstract \| Free Full Text" }	[ { "id": "2924", "date": "10 Jan 2014", "name": "Steven Niederer", "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 “JSim, an open-source modelling system for data analysis” provides a succinct update on the functionality and utility of the modelling platform JSim. The paper provides a concise description and link between the JSim community modelling philosophy and how this is facilitated by the JSim software platform. The complete description of the JSim environment will be of interest to the modelling community and this manuscript highlights much of the functionality that they require. This publication could be improved by addressing:As described in the article, the platform JSim has been developed over a period of over 40 years. Previous articles on JSim have been published and this article would be strengthened by highlighting the new features / functionality added to the platform since the previous JSim article. The article, as exemplified in the abstract, focuses on the technical functionality of JSim. If JSim can be readily set up or used by people who wish to analyse their experimental data with models, for example experimental researchers, as opposed to developing new models, then it would be worth highlighting this in the abstract and text. The article does not discuss or review alternate simulation platforms (for example COR, OpenCell, Continuity, CHASTE, SBML simulation environments). For new users wishing to make an informed decision it would be useful to highlight the differences between JSim and alternate platforms.Minor commentsIn the introduction, the statement that mathematical models provide clear and precise hypothesis that are susceptible to contradiction and that failure to fit leads to rejection needs to be more nuanced, particularly in the case of biology where comparisons are often made between deterministic models and variable experimental results. It is not clear in the loop section if JSim supports nested loops, this could be clarified. It would be of interest to provide some indicative performance measures. For example if simulating a cardiac action potential will JSim solve faster or slower than real time on a conventional desktop. The authors could comment on the utility or potential for adoption of new mark-up languages for spatial problems (FieldML) or problem definition formats (SED-ML). It would be interesting for the authors to comment on how or if they have verified the JSim code stack.", "responses": [ { "c_id": "776", "date": "17 Apr 2014", "name": "Erik Butterworth", "role": "Author Response", "response": "“As described in the article, the platform JSim has been developed over a period of over 40 years. Previous articles on JSim have been published and this article would be strengthened by highlighting the new features/functionality added to the platform since the previous JSim article.”This is the first \"complete\" of description of JSim to be published. Previous papers have made use of JSim as a tool, but did not describe the software or the systematic approach to it. A comparison of JSim functionality to its predecessors, SimCon and XSim, is in the section labeled \"Background\" in the paper's introduction. “The article, as exemplified in the abstract, focuses on the technical functionality of JSim. If JSim can be readily set up or used by people who wish to analyse their experimental data with models, for example experimental researchers, as opposed to developing new models, then it would be worth highlighting this in the abstract and text.”Good suggestion. One sentence is added to the abstract, and four sentences added on page 16 in a new section just before the Summary entitled \"Using archived models for analyzing one’s own data\" “The article does not discuss or review alternate simulation platforms (for example COR, OpenCell, Continuity, CHASTE, SBML simulation environments). For new users wishing to make an informed decision it would be useful to highlight the differences between JSim and alternative platforms.”Although this paper's main focus is a fairly detailed description of JSim, we've added a new section \"Some Alternative Simulation Platforms\" (page 18) that discusses alternative simulation platforms to JSim. Given the large number of available products (over 250 SBML-based products alone) the list is not exhaustive. Platforms discussed are Virtual Cell, COPASI, roadRunner, Chaste, PCenv, COR, OpenCell, OpenCOR and Continuity. The descriptions of these programs are not complete and there are no feature-to-feature comparison between them and JSim. Detailed comparison, evaluating available mathematical methods, features, portability, usability, performance, and scientific reproducibility would be a good idea for a follow-on paper.Minor Comments:“In the introduction, the statement that mathematical models provide clear and precise hypothesis that are susceptible to contradiction and that failure to fit leads to rejection needs to be more nuanced, particularly in the case of biology where comparisons are often made between deterministic models and variable experimental results.”The second paragraph of the Introduction has been revised accordingly. “It is not clear in the loop section if JSim supports nested loops, this could be clarified.”JSim currently supports 2 loop nesting levels (inner loops and outer loops). This has been clarified in the section \"Loops: Iterating solutions to exhibit behaviour\" (page 11). “It would be of interest to provide some indicative performance measures. For example if simulating a cardiac action potential will JSim solve faster or slower than real time on a conventional desktop.”Since JSim is a general purpose modeling engine, it can't be said that there is a particular \"standard\" model upon which to base comparisons. JSim users typically write models that run somewhere between a small fraction of a second to several minutes depending upon the complexity and numeric methods required. In particular, spatial models using PDEs compute much slower than compartmental models using ODEs. Most pharmacokinetic and whole body circulatory models run faster than real time, but there are exceptions, e.g. models with multistage receptors, transporters and complicated reaction networks. Under Run Time Performance on p 15, we've added a description of ODE and PDE assessment of model solution times in millisecond solution times per second of model time, comparing solvers and ODEs versus PDEs. This new section gives an overview of the various factors affecting run-time performance, and gives some run times for a selected set of models (including a BR cardiac action potential model). These run times vary from about 13,966 times faster than real time to about 5 times slower.We have a paper in preparation that discusses comparative performance in more detail, and will contrast JSim performance with several other computational engines for SBML and Matlab models. “ The authors could comment on the utility or potential for adoption of new mark-up languages for spatial problems (FieldML) or problem definition formats (SED-ML).”SED-ML support within JSim is currently under development. We have added a section entitled \"SED-ML Support\" to the fact \"Future Plans\" (page 20) section of the paper to describe our plans there.Support for FieldML is a more problematic. FieldML provides a mathematical framework for calculating a CellML point models over space, defined in terms of a real topological manifold. In the field of 2D and 3D computation, such manifold representations are often complex, irregular spatial grids. At present, JSim represents N-dimensional space as a cross product of regularly spaced grids. Support for generalized manifolds would be a major change in functionality requiring significant time and resources. While such a change might be contemplated some point in future, it is not within our immediate plans. JSim could support a limited version of FieldML, supporting the subset of real manifolds that are regular grids in N-space, but it is unclear what would be gained from a user's point of view. The primary advantage of standard ML support is the ability to access existing archives and exchange models between modeling systems. Unlike CellML and SBML, there is no evidence of large archives of FieldML models that can be run on a variety of platforms; our understanding is that FieldML-defined model can be run only in CMiss at this point. It can be considered bothersome to have separate archival forms for PDEs and ODEs. What one wants for ease of reproducibility is one self-consistent system, preferably all in one file. JSim does this for a limited repertoire of ODEs, 1- and 2-D PDEs, and DAEs. “It would be interesting for the authors to comment on how or if they have verified the JSim code stack.”Thanks for this suggestion. We've added a new section \"Code Verification\" (page 16) to address this topic. In brief, we use a combination of various strategies including comparison with analytic solutions, test statistics, comparison with other computational environments and round-tripping. Bug reports from modelers who find computations that violate physical intuition play a major role. An automated verification suite of anointed regression tests ensures continued compliance over subsequent releases." } ] }, { "id": "2922", "date": "23 Jan 2014", "name": "David Nickerson", "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 provides an introduction to, and description of, the JSim modelling system. The authors highlight the general purpose utility of this platform through the use of specific application examples, that are easily understood and followed by the reader and potential JSim users. This manuscript provides all appropriate links and examples that readers would require to get up and running with the JSim software.This paper is well written, with just a few points that the authors may want to consider in future revisions of the manuscript.The authors touch briefly on reuse of existing models/projects and the use of a CVS repository to archive the history of model development, as well as the discussion on modular modelling in the future developments section. The basis for this modularity and reuse seems to result in the development of a new, monolithic MML document for the assembled model. It might be useful to see if there are features in either JSim or MML that allow dynamic links to the source modules to be maintained allowing users to alternate sources or versions of the source modules (rather than the cut-and-paste style described in the manuscript). In addition to the versioning of the JSim input data (experimental data, MML, projects, etc.), it is often the case that a specific piece of work requires some minimum version of the software itself. I wonder if there is any link between project files and JSim releases? For example, are users browsing the Physiome Repository able to determine if they need to update their version of JSim prior to loading a project file (or if in fact the software handles this internally). The manuscript would benefit from a more thorough comparison of JSim to alternative tools, or at least some links to specific tools being contrasted in the article. There is no description of how spatial geometries (finite element meshes or finite difference grids) are defined in JSim. Are the evolving standards for such descriptions (e.g., FieldML or SBML-spatial) being used or are there plans to use such? A comparison with approaches taken by tools like the Virtual Cell or Chaste might be useful. The authors make no reference to the adoption or interchange with the SED-ML standard. It would be useful to discuss any plans in this regard. Similarly, the evolving COMBINE archive format has a large overlap in aims with the JSim project file and the authors might want to comment on any plans to make use of that archiving format or contributions in that direction. In some parts of the manuscript (e.g., the caption for figure 3) the description of the modelling/simulation example seems a bit excessive, and detracts from the primary focus of the article.Minor commentsPage 5, column 1, first paragraph: Antimony is mentioned as a model import source format, but that format is not defined previously. Page 10, column 1, third paragraph: \"...using several different numerical method,\" missing 's' on method.", "responses": [ { "c_id": "775", "date": "17 Apr 2014", "name": "Erik Butterworth", "role": "Author Response", "response": "“The authors touch briefly on reuse of existing models/projects and the use of a CVS repository to archive the history of model development, as well as the discussion on modular modelling in the future developments section. The basis for this modularity and reuse seems to result in the development of a new, monolithic MML document for the assembled model. It might be useful to see if there are features in either JSim or MML that allow dynamic links to the source modules to be maintained allowing users to alternate sources or versions of the source modules (rather than the cut-and-paste style described in the manuscript).”We agree that JSim and MML modularity could be improved. We are currently in the design stages of a more modular form of MML that moves away from the current monolithic approach. Design goals for the new MML include the following: the ability to draw upon and reuse code modules stored either in a project file or in a permanent archived location on-line; support for the recently developed SBML \"Hierarchical Model Composition\" package; support for support for CellML v1.1 modular structuring; support for structured multiple reuse on modules (e.g. support of multiple parallel pathways); support for run-time switching of alternative modules; compatibility with existing MML models. However, we feel this work is not yet far enough along for public presentation. “ In addition to the versioning of the JSim input data (experimental data, MML, projects, etc.), it is often the case that a specific piece of work requires some minimum version of the software itself. I wonder if there is any link between project files and JSim releases? For example, are users browsing the Physiome Repository able to determine if they need to update their version of JSim prior to loading a project file (or if in fact the software handles this internally).”All NSR provided models are curated to work properly with the current version of JSim, which is the version run in live Web applets. Users may also download JSim project for use on their own workstation, which may have an older JSim version installed. The vast majority of the time, JSim runs correctly on projects created by different JSim versions. The only difference will be that, when running an older JSim version, the latest features will not be available. There are three primary reasons why this is the case. First, the basic functionality of JSim and project file structure has been stable for many years. Second, the XML structure of project files allows old versions of JSim to ignore features in newer project files. Third, a version number tag in each project file allows newer versions of JSim to migrate old projects to the new format. Occasionally, an NSR modeler discovers a JSim bug that must be fixed for proper operation. Release of that model will be delayed until a new version of JSim is released. In that case, our best practices recommendation is to note the minimum JSim version required to run the model on the relevant web page. For published papers, where it is important to reproduce figures \"warts and all\", it is recommended that the papers list the version of JSim used to produce the results. Based on your comment, we discussed the possibility of making the JSim project version for each model more visible on our site. The consensus was that, given the generally high level of compatibility, this information would be mostly misleading to users by assigning anundeserved level of attention to the version number. “The manuscript would benefit from a more thorough comparison of JSim to alternative tools, or at least some links to specific tools being contrasted in the article.”We've added a new section \"Some Alternative Simulation Platforms\" (page 18) that briefly discusses some alternative simulation platforms to JSim. Given the large number of available products (over 250 SBML-based products alone) the list is not exhaustive. Platforms mentioned are Virtual Cell, COPASI, roadRunner, Chaste, PCenv, COR, OpenCell, OpenCOR and Continuity. Given limited space, the descriptions of these programs are not complete nor is there a feature-to-feature comparison with JSim. Such a comparison, evaluating available mathematical methods, features, portability, usability, performance and scientific reproducibility would be a good idea for a follow-on paper tothis one. “There is no description of how spatial geometries (finite element meshes or finite difference grids) are defined in JSim. Are the evolving standards for such descriptions (e.g., FieldML or SBML-spatial) being used or are there plans to use such? A comparison with approaches taken by tools like the Virtual Cell or Chaste might be useful.”In MML, at present, spatial grids are defined in regular N-space. The 2D and 3D geometries used in finite element methods and higher dimensional PDEs are often specified using much more complex data structures. We forsee adding support for 2D and 3D PDEs on regularly spaced grids to JSim at some point in the not-too-distant future (JSim's MML compiler already supports parsing 2D and 3D PDEs; what is currently missing is appropriate numeric libraries). However, support for irregular high-dimension grids would require major changes to JSim and is not envisioned in the near future. We have monitored the standardization efforts of the SBML spatial extension and FieldML. At the point when any serious body of models exists for either of these formats, we will consider adding support for them to JSim's existing SBML and CellML translators. The utility of this effort will depend upon how many models in that body are runnable given JSim's limitation to regularly spaced grids. “The authors make no reference to the adoption or interchange with the SED-ML standard. It would be useful to discuss any plans in this regard. Similarly, the evolving COMBINE archive format has a large overlap in aims with the JSim project file and the authors might want to comment on any plans to make use of that archiving format or contributions in that direction.”SED-ML support within JSim is currently under development. It is described in a section entitled \"SED-ML Support\" to the \"Future Developments\" section (page 20) of the paper to describe our plans there.COMBINE looks like an interesting project, albeit one in the preliminary stages. But why would it be advantageous to migrate away from JSim project files? Should COMBINE become widely adopted, we will certainly consider supporting COMBINE import and export options within JSim. I think COMBINE is too preliminary for mention in the paper, however. “In some parts of the manuscript (e.g., the caption for figure 3) the description of the modelling/simulation example seems a bit excessive, and detracts from the primary focus of the article.”Figure 3 is problem, we agree. But it takes quite a description to convey understudying of what can be rapidly learned from such multidimensional plots. Complex, but easy to produce and rapid to interpret after learning how. We think the legend is needed so that a reader can determine how valuable the feature might be, even though it takes space.Minor comments:“Page 5, column 1, first paragraph: Antimony is mentioned as a model import source format, but that format is not defined previously.”“ Page 10, column 1, third paragraph: \"...using several different numerical method,\" missing 's' on method.”Thank you. Text now corrected." } ] } ]	1	https://f1000research.com/articles/2-288
https://f1000research.com/articles/3-137/v1	27 Jun 14	{ "type": "Research Article", "title": "How do eubacterial organisms manage aggregation-prone proteome?", "authors": [ "Rishi Das Roy", "Manju Bhardwaj", "Vasudha Bhatnagar", "Kausik Chakraborty", "Debasis Dash", "Rishi Das Roy", "Manju Bhardwaj", "Vasudha Bhatnagar" ], "abstract": "Eubacterial genomes vary considerably in their nucleotide composition. The percentage of genetic material constituted by guanosine and cytosine (GC) nucleotides ranges from 20% to 70%. It has been posited that GC-poor organisms are more dependent on protein folding machinery. Previous studies have ascribed this to the accumulation of mildly deleterious mutations in these organisms due to population bottlenecks. This phenomenon has been supported by protein folding simulations, which showed that proteins encoded by GC-poor organisms are more prone to aggregation than proteins encoded by GC-rich organisms. To test this proposition using a genome-wide approach, we classified different eubacterial proteomes in terms of their aggregation propensity and chaperone-dependence using multiple machine learning models. In contrast to the expected decrease in protein aggregation with an increase in GC richness, we found that the aggregation propensity of proteomes increases with GC content. A similar and even more significant correlation was obtained with the GroEL-dependence of proteomes: GC-poor proteomes have evolved to be less dependent on GroEL than GC-rich proteomes. We thus propose that a decrease in eubacterial GC content may have been selected in organisms facing proteostasis problems.", "keywords": [ "Eubacterial organisms have genomes that vary largely in their nucleotide compositions. In this kingdom", "the GC content varies from 20% to 70% of the genome and this large variation has been documented in a number of reports that have aimed to explain it1–3. The amino acid compositions are also different in eubacterial proteomes due to the variation of GC content4. It has been reported that these difference of amino acid compositions alter the characteristics of proteomes and as a consequence", "proteins of GC-poor genomes are more prone to misfolding and aggregation compare to GC-rich genomes5", "6. It has been hypothesized that GroEL plays a major role", "if not an essential role", "in the evolution of GC-poor organisms by buffering deleterious mutations that are fixed due to population bottlenecks7–9. This has been supported by the observation that many of the small GC-poor endosymbionts tend to overexpress GroEL10–12." ], "content": "Introduction\n\nEubacterial organisms have genomes that vary largely in their nucleotide compositions. In this kingdom, the GC content varies from 20% to 70% of the genome and this large variation has been documented in a number of reports that have aimed to explain it1–3. The amino acid compositions are also different in eubacterial proteomes due to the variation of GC content4. It has been reported that these difference of amino acid compositions alter the characteristics of proteomes and as a consequence, proteins of GC-poor genomes are more prone to misfolding and aggregation compare to GC-rich genomes5,6. It has been hypothesized that GroEL plays a major role, if not an essential role, in the evolution of GC-poor organisms by buffering deleterious mutations that are fixed due to population bottlenecks7–9. This has been supported by the observation that many of the small GC-poor endosymbionts tend to overexpress GroEL10–12.\n\nHowever, the proposed chaperone dependence of GC-poor organisms does not explain why some of the GC-poor endosymbionts of the mycoplasma group have lost the groEL copy from their genome13. It is notable that these are the only known eubacterial organisms to have lost this gene. This observation led us to test the proposed relationship of GC poorness of genome with the aggregation propensity of the encoded proteome.\n\nObtaining information on the aggregation propensity of proteins from different organisms is a challenging task. However, there has already been a careful characterization of the aggregation propensity of different Escherichia coli proteins that was conducted in a high-throughput manner14–16. Kerner et al. classified the GroEL substrates into Class I, II or III based on the interaction strength and on the stringency of their requirement for GroEL. Class III (C3) substrates were completely dependent on GroEL for folding, whereas Class II (C2) substrates were partially dependent. Class I (C1) proteins interacted weakly with GroEL and were able to fold spontaneously. In a trivial approach, homologs of GroEL-dependent proteins may be identified in other organisms13,17. This approach however fails to predict the evolution of protein dependence on GroEL correctly, as the sequence differences between species have the potential to introduce or remove kinetic traps from folding pathways, thereby altering their dependence on GroEL. In addition to the solubility of the E. coli proteome in a chaperone-free system, substrates of another chaperone DnaK were also identified by two independent research groups18,19. Applications developed primarily on machine learning algorithms to classify soluble or GroEL substrates16,18,20–24 are already available. However, these classifiers have not been trained with curated data prepared from multiple experimental results14,15,18,19. In this study, we have constructed a more reliable training dataset to build classifiers to determine the aggregation propensity and GroEL dependency in 1132 eubacterial proteomes, based solely on the amino acid sequences. We show a distinct trend in the aggregation propensity of proteins of an organism in relation to the GC content. Surprisingly, aggregation propensity decreased with lower GC content independent of symbiotic characteristics, suggesting that GC-poor organisms have indeed evolved a proteome that is devoid of aggregation-prone proteins.\n\n\nMaterials and methods\n\nThe aggregation-prone proteins of the eSOL database18,25 are dependent on the chaperone network of E. coli to get their three dimensional native structure. GroEL and DnaK are two important components of this network and their substrates have been extensively studied via different experimental methods14,15,19,26. The integration of all the available information reveals that about half (457) of the soluble or chaperone-independent proteins identified by Niwa et al. were found to be GroEL- or DnaK-dependent18 (Figure 1). To construct a more reliable training set, we removed these proteins from the soluble set. Thus, proteins identified as chaperone-dependent by more than one study, were only considered as aggregation-prone proteins. Furthermore, the proteins which were more than 30% (amino acid) sequence similarity among the remaining proteins were removed using CD-HIT27 clustering program. Therefore the final training set comprised of 502 aggregation prone and 475 soluble proteins.\n\nA Venn diagram of proteins of E. coli identified by different experimental studies shows that ~45% of soluble proteins reported by Niwa et al. overlap with GroEL/S or DnaK substrates (soluble proteins are defined as having solubility >70% and aggregation-prone proteins have solubility <30%).\n\nThe classifiers in this study were built with Pro-Gyan28 software. Pro-Gyan builds classifiers directly from training data set given in FASTA format by selecting relevant features from a large number of unbiased features. Following metrics which are useful to evaluate performance of machine learning classifiers were reported by Pro-Gyan.\n\nAccuracy(Acc)= (TP + TN)/(TP + TN + FP + FN)\n\nSensitivity or Recall (Sn) = TP/(TP + FN)\n\nSencificity (Sp) = TN/(FP + TN)\n\nMatthews correlation coefficient (MCC) = (TPTN–FPFN)/{(TP+FP)(TN+FN)(TP+FN)(TN+FP)}\n\nwhere TP = True Positive, TN = True Negative, FP = False Positive, FN = False Negative predicted by the classifier.\n\nAdditionally, receiver operating characteristic (ROC) curves and area under this curve (AUC)29 were also generated.\n\nThe protein sequences of microbial genomes were downloaded from the Microbial Genome Database30 (archive no. mbgd_2011-01). To identify the chaperonins in the microbial organisms, chaperonin homologs were searched for using BLAST (e-value 110-4) against a chaperonin database cpnDB31 downloaded on June 2011. The 16S rRNA nucleotide sequence of E. coli was acquired from SILVA32 and homologous were searched for in other microbial organisms using BLAST (e-value 110-4). GC contents for microbial genomes were calculated using following equation\n\nGC content = (G + C)/(total bases),\n\nwhere G = number of guanosine and C = number of cytosine.\n\nThe Kendall correlation and analysis of covariance were performed in R33 statistical computing environment using the package ‘stats’ version 2.15.3. To account the effect of evolution on different traits of bacterial genomes, we performed phylogenetic independent contrast through the PDAP34 module on Mesquite35 application.\n\n\nResults and discussion\n\nRecently protein solubility has been carefully measured in a chaperone-free system and the information has been made available through the eSol database18. Few classification models developed on this database can segregate soluble proteins from chaperone-dependent proteins22–24. However, these web-based classifiers are not suitable to classify large numbers of proteomes, and their soluble or negative training dataset (proteins not aggregation-prone or soluble) are not carefully curated, as most of the soluble proteins from eSol database are substrates of DnaK19 or GroEL14,15 (Figure 1). Therefore we built a classifier containing a curated list of aggregation-prone proteins and soluble proteins. The classifier was built using Pro-Gyan28 which generates 5038 different features from a set of class labelled protein sequences and selects the “maximum relevant minimum redundant” feature subset. Finally, the tool built a support vector machine (SVM)36 classifier by five-fold cross validation. The classifier attained an accuracy of 83.21% with 0.66 MCC (Table 1). Although Pro-Gyan generated classifier was trained with a rigorously curated training data set, it performs equivalent to Fang et al.’s classifier and better than others22–24. The receiver operating characteristic (ROC) curves of the classifier are shown in Figure 2. For interested users, the classifier is available in ZENODO (https://zenodo.org/record/10442).\n\n Built on a curated training data set.\n\nROC curves of the soluble protein classifier (SolubEcoli.pgc) and the GroEL obligate protein classifier (GDP1.pgc). The areas under the curves (AUC) are given in the legend.\n\nTo build the classifier, Pro-Gyan28 selected 24 relevant features through an automated process. The top ten significant (by Mann-Whitney test) features were the sequence patterns, the pseudo amino acid composition37 of phenylalanine (F), aspartic (D) and glutamic (E) acid, the distribution of positively charged amino acids, the features calculated from FoldIndex38 and the auto-correlation of hydrophobicity and relative mutability (Table 2). The remaining selected features (Table 2) were enriched with auto-correlation measurement of amino acid indices such as steric parameter, free energy, accessible surface area, polarizability, residue volume etc. The features which represent patterns of physico-chemical properties encrypted in protein sequences were unique to SolubEcoli.pgc when compared to earlier methods.\n\n†Internal feature id of the Pro-Gyan application.\n\nFrom the analysis of features, it was noticed that the compositions of amino acids are significantly different within aggregation prone and soluble proteins. Sequence features of amino acids have been used to understand protein overexpression related to toxicity39. Additionally, it has been also shown that the amino acid composition is drastically altered in organisms with GC-poor genomes4,40. There are multiple amino acids that change in frequency as a function of GC content (Figure 3) and this change that has been attributed to the difference in the GC content in the codons of these amino acids. On the basis of these differences, it has been reported that proteins encoded by GC-poor organisms should be more prone to aggregation than proteins encoded by GC-rich organisms5,6. However, the GC composition of the training data showed that aggregation-prone proteins were significantly more GC-rich than the soluble proteins (Figure 4, Mann-Whitney test p-value = 1.3e-15). Subsequently, we sought to verify the fraction of aggregation-prone proteins across different bacterial proteomes. We used the SolubEcoli.pgc classifier to predict aggregation-prone proteins in 1132 eubacterial species. Our prediction on bacterial genomes showed that the fAg (aggregation prone proteins as fraction of proteome) of a genome correlates positively with the GC composition (Kendall tau=0.38 p-value < 2.2e-16) (Figure 5A). We further examined the correlation, with respect to phylogenetic ancestry, using the Mesquite system35, because the Kendall correlation assumes that observations are independent even if organisms are linked through common ancestors41. The required phylogenetic tree was constructed from the 16S rRNA gene sequences of 570 bacteria42. We found a significant correlation (0.4, p-value < 2.2e-16) between independent contrasts of GC content and fAG (Figure 5B). This corroborated well with the difference seen between soluble and aggregation-prone proteins in E. coli (Figure 4). Thus the increase in the GC composition of a genome may encode proteome that harbours a higher fraction of aggregation-prone proteins.\n\nThe x-axis of each subplot shows for GC composition of each genome whereas y-axis shows corresponding amino acid composition.\n\nIn E. coli, aggregation-prone proteins contain higher GC-content than soluble proteins. Mann-Whitney test p-value (*) is 1.3e-15.\n\n(A) GC content of the genome correlates with the fraction of proteome that is aggregation-prone (fAg) (analysis of 570 bacterial genomes using the classifier). Rank-based correlation is provided along with the p-value. The black line shows a linear regression model. (B) The relationship between GC content and fAg was obtained through a phylogenetically independent contrast method (570 bacteria). A positive correlation (0.4) was identified between GC content and fAg (p-value < 2.1e-16).\n\nThis is in contrast to previous reports hypothesizing that GC-poor organisms have unstable and aggregation prone proteomes. Notably, the earlier hypothesis that GC poorness is associated with GroEL-dependent aggregation-prone proteomes was based on the observation that GroEL is overexpressed in GC-poor organisms. Therefore, to segregate GroEL-dependent proteins from aggregation-prone proteomes, we developed another classifier (ZENODO, https://zenodo.org/record/10442/) trained with 475 curated soluble and 83 GroEL obligate (Class 3 or C3) proteins14. The classifier achieved an accuracy of 92.29% with MCC of 0.69. We used GDP1.pgc to identify the C3 proteins within aggregation-prone proteins (predicted by SolubEcoli.pgc) to examine the evolution of the GroEL-dependent proteome with GC composition. Indeed we found that the fC3 (fraction of C3 proteins) of bacterial proteome are more correlated with GC content than the fAg fraction (Figure 6A). The phylogenetically independent contrasts of fC3 and GC also correlated strongly (0.7, p-value < 2.2e-16, Figure 6B). The phylum Tenericutes, members of which have GC-poor genomes, was predicted to encode less GroEL-dependent proteins. Mycoplasma and Ureaplasma are the main genera of the phylum Tenericutes and many species of these groups lack GroEL43. In our analysis, we also observed that the Tenericutes without GroEL (red dots in Figure 6A) had very few fC3 proteins. This motivated us to investigate the effect of groEL copy number on misfolded proteins. Interestingly, there was a strong correlation between the groEL copy number and the fraction of genome coding for C3 proteins (Figure 6C). Due to the presence of noise in the experimental data, we tried to benchmark the classifiers. Fujiwara et al. reported that five C3 homologs of groEL-lacking Ureaplasma urealyticum are soluble in GroEL depleted cells26. Hence, we also examined the tolerance of our classifiers by predicting the GroEL dependency of these homologs. Four of these homologs were predicted to be GroEL independent with a high confidence score (Table 3). Overall, the results indicated that C3 proteins and in general aggregation-prone proteins do decrease with the GC content of genomes.\n\n(A) Correlation of GC content with the fraction of the proteome that is GroEL obligate (fC3) over 570 bacterial genomes. Members of the phylum Tenericutes with and without the groEL gene are coloured in blue and red, respectively. Rank-based correlation is provided along with p-value. The black line shows a logarithmic regression model. (B) A positive correlation (0.7) was identified between independent contrast of GC content and fC3 with respect to phylogenetic information of bacterial genomes (570 bacteria, p-value < 2.2e-16). (C) The organisms were classified based on the number of groEL genes present in the genome. fC3 exhibited a significant increase with an increase in the number of genome-encoded groEL copies. The p-values were calculated by Mann-Whitney test using two-sided hypothesis.\n\nThe homologous were found in U. urealyticum by NCBI BLAST at a threshold of E value of 1e45. Then the aggregation propensity and GroEL dependency of these proteins were classified by SolubEcoli.pgc and GDP1.pgc.\n\nEndosymbionts are crucial to this study as the literature suggests that these organisms have undergone bottlenecks during evolution44. It is hypothesized that these organisms have accumulated more deleterious mutations compared to non-endosymbionts8. If this were true then endosymbionts should show a greater aggregation propensity or dependence on GroEL than that predicted by the GC content of free-living eubacterial species. To measure the impact of a symbiotic relationship on C3 proteins, we performed an analysis of covariance ANCOVA on 570 eubacterial species42. There was no significant effect of a symbiotic relationship on fAG/fC3 (p-value 0.24/0.65, Data set) or significant interaction (p-value=0.36/0.38) with GC composition (Figure 7). Thus we were unable to obtain proof for any association of a bottleneck in evolutionary history with protein aggregation propensity. Therefore we rule out the possibility of bottleneck evolution as the reason for the evolution of GroEL-independent proteomes like Ureaplasma and GroEL-independent mycoplasma species.\n\nThe ANCOVA test on 570 organisms showed that a symbiotic relationship has no significant effect or interaction with GC content on the aggregation propensity or GroEL-dependency of the proteins of an organism.\n\n\nConclusions\n\nSeveral machine learning (ML) classifiers have been developed to predict aggregation-prone or GroEL-dependent proteins, but very few of them used data sets generated and curated from multiple experimental studies. Our classifiers were based on curated data from multiple studies and performed well also against the false positive C3 homologs of Ureaplasma, showing accuracy and noise tolerance. According to previous theories, GC-poor organisms might have evolved through population bottlenecks. This allows mildly deleterious mutations to be fixed in the population with a high probability2,44. It has been hypothesized that the GC-poor genomes that accumulated a large number of deleterious mutations in the course of evolution, through population bottlenecks and hence harbour proteins that are aggregation-prone. Although overexpressions of chaperones are observed in GC-poor organisms that have reduced genomes, there are also other GC-poor organisms that lack GroEL. Our work provides strong evidence that the general stability of the proteome increases with the decrease in GC content of eubacterial genomes. Decrease in GC content restricts the amino acid space that the organism can sample, thereby compromising protein evolution. We hypothesise that, even with this limited amino acid space, GC-poor organisms are still abundant as growth is facilitated under conditions that compromise protein folding capacity. This antagonism between ability to evolve and folding advantage could be crucial in facilitating protein evolution in the presence of chaperones and other folding machineries45–48.\n\nOur work suggests that organisms facing continuous proteostasis problems would tend to shift towards a more GC-poor genome. This is supported by findings of Xia et al.49 who have reported that the preponderance of GC to AT conversions during high temperature laboratory adaptation of Pasteurella multocida. Further in vitro evolution experiments will be required to demonstrate whether laboratory adaptation to low GC content may provide folding advantage.\n\n\nData availability\n\nF1000Research: Dataset 1. Application of SolubEcoli.pgc and GDP1.pgc classifiers, 10.5256/f1000research.4307.d2962455.\n\nZENODO: Training data of protein classifier SolubEcoli.pgc and GDP1.pgc, doi: 10.5281/zenodo.1044256.", "appendix": "Author contributions\n\n\n\nKC and DD designed the study. RDR carried out all the analysis. MB and VB participated in the design and discussion of the study. RDR, KC and DD prepared the manuscript and all authors agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nRDR acknowledges HP and CSIR for funding the study. DD and KC acknowledge CSIR for funding through IGIB. KC acknowledges Wellcome Trust – DBT India alliance for funding and fellowship.\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\nNishida H: Evolution of genome base composition and genome size in bacteria. Front Microbiol. 2012; 3: 420. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCutcheon JP, Moran NA: Extreme genome reduction in symbiotic bacteria. Nat Rev Microbiol. 2012; 10(1): 13–26. PubMed Abstract \| Publisher Full Text\n\nGuo FB, Lin H, Huang J: A plot of G + C content against sequence length of 640 bacterial chromosomes shows the points are widely scattered in the upper triangular area. Chromosome Res. 2009; 17(3): 359–364. PubMed Abstract \| Publisher Full Text\n\nLightfield J, Fram NR, Ely B: Across bacterial phyla, distantly-related genomes with similar genomic GC content have similar patterns of amino acid usage. PLoS One. 2011; 6(3): e17677. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Ham RC, Kamerbeek J, Palacios C, et al.: Reductive genome evolution in Buchnera aphidicola. Proc Natl Acad Sci U S A. 2003; 100(2): 581–586. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBastolla U, Moya A, Viguera E, et al.: Genomic determinants of protein folding thermodynamics in prokaryotic organisms. J Mol Biol. 2004; 343(5): 1451–1466. PubMed Abstract \| Publisher Full Text\n\nFares MA, Moya A, Barrio E: GroEL and the maintenance of bacterial endosymbiosis. Trends Genet. 2004; 20(9): 413–416. PubMed Abstract \| Publisher Full Text\n\nFares MA, Ruiz-Gonzalez MX, Moya A, et al.: Endosymbiotic bacteria: groEL buffers against deleterious mutations. Nature. 2002; 417(6887): 398. PubMed Abstract \| Publisher Full Text\n\nMoran NA: Accelerated evolution and Muller’s rachet in endosymbiotic bacteria. Proc Natl Acad Sci U S A. 1996; 93(7): 2873–2878. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAksoy S: Molecular analysis of the endosymbionts of tsetse flies: 16S rDNA locus and over-expression of a chaperonin. Insect Mol Biol. 1995; 4(1): 23–29. PubMed Abstract\n\nClark MA, Baumann L, Baumann P: Sequence analysis of a 34.7–kb DNA segment from the genome of Buchnera aphidicola (endosymbiont of aphids) containing groEL, dnaA, the atp operon, gidA, and rho. Curr Microbiol. 1998; 36(3): 158–163. PubMed Abstract \| Publisher Full Text\n\nWilcox JL, Dunbar HE, Wolfinger RD, et al.: Consequences of reductive evolution for gene expression in an obligate endosymbiont. Mol Microbiol. 2003; 48(6): 1491–1500. PubMed Abstract \| Publisher Full Text\n\nWilliams TA, Fares MA: The effect of chaperonin buffering on protein evolution. Genome Biol Evol. 2010; 2: 609–619. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKerner MJ, Naylor DJ, Ishihama Y, et al.: Proteome-wide analysis of chaperonin-dependent protein folding in Escherichia coli. Cell. 2005; 122(2): 209–220. PubMed Abstract \| Publisher Full Text\n\nChapman E, Farr GW, Usaite R, et al.: Global aggregation of newly translated proteins in an Escherichia coli strain deficient of the chaperonin GroEL. Proc Natl Acad Sci U S A. 2006; 103(43): 15800–15805. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRaineri E, Ribeca P, Serrano L, et al.: A more precise characterization of chaperonin substrates. Bioinformatics. 2010; 26(14): 1685–1689. PubMed Abstract \| Publisher Full Text\n\nBogumil D, Dagan T: Chaperonin-dependent accelerated substitution rates in prokaryotes. Genome Biol Evol. 2010; 2: 602–608. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNiwa T, Kanamori T, Ueda T, et al.: Global analysis of chaperone effects using a reconstituted cell-free translation system. Proc Natl Acad Sci U S A. 2012; 109(23): 8937–8942. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalloni G, Chen T, Schermann SM, et al.: DnaK Functions as a Central Hub in the E. coli Chaperone Network. Cell Rep. 2012; 1(3): 251–264. PubMed Abstract \| Publisher Full Text\n\nTartaglia GG, Dobson CM, Hartl FU, et al.: Physicochemical determinants of chaperone requirements. J Mol Biol. 2010; 400(3): 579–588. PubMed Abstract \| Publisher Full Text\n\nNoivirt-Brik O, Unger R, Horovitz A: Low folding propensity and high translation efficiency distinguish in vivo substrates of GroEL from other Escherichia coli proteins. Bioinformatics. 2007; 23(24): 3276–3279. PubMed Abstract \| Publisher Full Text\n\nFang Y, Fang J: Discrimination of soluble and aggregation-prone proteins based on sequence information. Mol BioSyst. 2013; 9(4): 806–811. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStiglic G, Kocbek S, Pernek I, et al.: Comprehensive decision tree models in bioinformatics. PLoS One. 2012; 7(3): e33812. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKlus P, Bolognesi B, Agostini F, et al.: The cleverSuite Approach for Protein Characterization: Predictions of Structural Properties, Solubility, Chaperone Requirements and RNA-Binding Abilities. Bioinformatics. 2014; 30(11): 1601–1608. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNiwa T, Ying BW, Saito K, et al.: Bimodal protein solubility distribution revealed by an aggregation analysis of the entire ensemble of Escherichia coli proteins. Proc Natl Acad Sci U S A. 2009; 106(11): 4201–4206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFujiwara K, Ishihama Y, Nakahigashi K, et al.: A systematic survey of in vivo obligate chaperonin-dependent substrates. EMBO J. 2010; 29(9): 1552–1564. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi W, Godzik A: Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006; 22(13): 1658–1659. PubMed Abstract \| Publisher Full Text\n\nDas Roy R, Dash D: Selection of relevant features from amino acids enables development of robust classifiers. Amino Acids. 2014; 46(5): 1343–1351. PubMed Abstract \| Publisher Full Text\n\nZweig MH, Campbell G: Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem. 1993; 39(4): 561–577. PubMed Abstract\n\nUchiyama I, Higuchi T, Kawai M: MBGD update 2010: toward a comprehensive resource for exploring microbial genome diversity. Nucleic Acids Res. 2010; 38(Database issue): D361–D365. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHill JE, Penny SL, Crowell KG, et al.: cpnDB: a chaperonin sequence database. Genome Res. 2004; 14(8): 1669–1675. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPruesse E, Quast C, Knittel K, et al.: SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res. 2007; 35(21): 7188–7196. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTeam RC: R: A language and environment for statistical computing. R foundation for Statistical Computing. 2005. Reference Source\n\nMidford P, Garland T Jr, Maddison W: PDAP Package of Mesquite. Version 1.14. 2008. Reference Source\n\nMesquite: a modular system for evolutionary analysis. Version 2.75. Reference Source\n\nVapnik V: The nature of statistical learning theory. 2nd edition. Springer; 1999. Reference Source\n\nChou KC: Prediction of protein cellular attributes using pseudo-amino acid composition. Proteins. 2001; 43(3): 246–255. PubMed Abstract \| Publisher Full Text\n\nPrilusky J, Felder CE, Zeev-Ben-Mordehai T, et al.: FoldIndex©: a simple tool to predict whether a given protein sequence is intrinsically unfolded. Bioinformatics. 2005; 21(16): 3435–3438. PubMed Abstract \| Publisher Full Text\n\nSingh GP, Dash D: Electrostatic mis-interactions cause overexpression toxicity of proteins in E. coli. PLoS One. 2013; 8(5): e64893. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBohlin J, Brynildsrud O, Vesth T, et al.: Amino acid usage Is asymmetrically biased in AT-and GC-rich microbial genomes. PLoS One. 2013; 8(7): e69878. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarland T, Bennett AF, Rezende EL: Phylogenetic approaches in comparative physiology. J Exp Biol. 2005; 208(Pt 16): 3015–3035. PubMed Abstract \| Publisher Full Text\n\nMazurie A, Bonchev D, Schwikowski B, et al.: Evolution of metabolic network organization. BMC Syst Biol. 2010; 4: 59. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClark GW, Tillier ER: Loss and gain of GroEL in the Mollicutes. Biochem Cell Biol. 2010; 88(2): 185–194. PubMed Abstract \| Publisher Full Text\n\nMira A, Moran NA: Estimating population size and transmission bottlenecks in maternally transmitted endosymbiotic bacteria. Microb Ecol. 2002; 44(2): 137–143. PubMed Abstract \| Publisher Full Text\n\nBandyopadhyay A, Saxena K, Kasturia N, et al.: Chemical chaperones assist intracellular folding to buffer mutational variations. Nat Chem Biol. 2012; 8(3): 238–245. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRutherford SL, Lindquist S: Hsp90 as a capacitor for morphological evolution. Nature. 1998; 396(6709): 336–342. PubMed Abstract \| Publisher Full Text\n\nQueitsch C, Sangster TA, Lindquist S: Hsp90 as a capacitor of phenotypic variation. Nature. 2002; 417(6889): 618–624. PubMed Abstract \| Publisher Full Text\n\nRohner N, Jarosz DF, Kowalko JE, et al.: Cryptic variation in morphological evolution: HSP90 as a capacitor for loss of eyes in cavefish. Science. 2013; 342(6164): 1372–1375. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXia X, Wei T, Xie Z, et al.: Genomic changes in nucleotide and dinucleotide frequencies in Pasteurella multocida cultured under high temperature. Genetics. 2002; 161(4): 1385–1394. PubMed Abstract\n\nChou KC: Prediction of protein subcellular locations by incorporating quasi-sequence-order effect. Biochem Biophys Res Commun. 2000; 278(2): 477–483. PubMed Abstract \| Publisher Full Text\n\nBum Ju L, Keun Ho R: Feature Extraction from Protein Sequences and Classification of Enzyme Function. In BioMedical Engineering and Informatics, 2008 BMEI 2008 International Conference on; 27–30 May 2008. 2008; 138–142. Publisher Full Text\n\nShamim MT, Anwaruddin M, Nagarajaram HA: Support Vector Machine-based classification of protein folds using the structural properties of amino acid residues and amino acid residue pairs. Bioinformatics. 2007; 23(24): 3320–3327. PubMed Abstract \| Publisher Full Text\n\nLi ZR, Lin HH, Han LY, et al.: PROFEAT: a web server for computing structural and physicochemical features of proteins and peptides from amino acid sequence. Nucleic Acids Res. 2006; 34(Web Server issue): W32–W37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDubchak I, Muchnik I, Holbrook SR, et al.: Prediction of protein folding class using global description of amino acid sequence. Proc Natl Acad Sci U S A. 1995; 92(19): 8700–8704. PubMed Abstract \| Free Full Text\n\nDas Roy R, Bhardwaj M, Bhatnagar V, et al.: Application of SolubEcoli.pgc and GDP1.pgc classifiers. F1000Research. Data Source\n\nDas Roy R, Chakraborty K, Das D: Training data of protein classifier SolubEcoli.pgc and GDP1.pgc. 2014. Data Source" }	[ { "id": "5273", "date": "15 Jul 2014", "name": "Amnon Horovitz", "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 study, the authors describe machine-learning classifiers for predicting aggregation propensities of proteins. A novel aspect of this work is that the classifiers are based on experimental data obtained from different sources regarding chaperone dependence (GroEL or DnaK) and solubility in chaperone-free systems. The authors then use their machine-learning approach to test for ~1100 eubacterial proteomes whether a low GC content correlates with a greater tendency to aggregate or mis-fold as suggested by earlier studies. One possibility is that the GC content affects the amino acid composition of proteins that, in turn, affects their folding and aggregation propensities. The authors show in Figure 3 of the paper that amino acid compositions do indeed correlate strongly with GC content (the Figure shows data for all 20 naturally occurring amino acids, although its title suggests otherwise). They then show that aggregation-prone proteins have a higher GC content than soluble proteins. This finding is not really new since it has already been reported (see ref. 21 in the paper) that GroEL substrates tend to have a relatively high GC content. However, the observation for a large number of genomes that the fraction of aggregation-prone proteins increases with an increasing GC content is novel. A weakness of the paper is that the authors do not discuss the problems of distinguishing between (i) causation and correlation; and (ii) cause and effect. In other words, is the correlation between high GC content and an increased tendency to mis-fold due to some unidentified factor(s) that correlates with both GC content and mis-folding or to a direct effect? In addition, it is also possible (at least in principle) that a high GC content reflects selection against mis-folding rather than being one of its ‘causes’.", "responses": [] }, { "id": "6283", "date": "01 Oct 2014", "name": "Annalisa Pastore", "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 genesis of this paper is the proposal that genomes containing a poor percentage of guanosine and cytosine (GC) nucleotide pairs lead to proteomes more prone to aggregation than those encoded by GC-rich genomes. As a consequence these organisms are also more dependent on the protein folding machinery. If true, this interesting hypothesis could establish a direct link between the tendency to aggregate and the genomic code.In their paper, the authors have tested the hypothesis on the genomes of eubacteria using a genome-wide approach based on multiple machine learning models. Eubacteria are an interesting set of organisms which have an appreciably high variation in their nucleotide composition with the percentage of CG genetic material ranging from 20% to 70%. The authors classified different eubacterial proteomes in terms of their aggregation propensity and chaperone-dependence. For this purpose, new classifiers had to be developed which were based on carefully curated data. They took account for twenty-four different features among which are sequence patterns, the pseudo amino acid composition of phenylalanine, aspartic and glutamic acid, the distribution of positively charged amino acids, the FoldIndex score and the hydrophobicity. These classifiers seem to be altogether more accurate and robust than previous such parameters.The authors found that, contrary to what expected from the working hypothesis, which would predict a decrease in protein aggregation with an increase in GC richness, the aggregation propensity of proteomes increases with the GC content and thus the stability of the proteome against aggregation increases with the decrease in GC content. The work also established a direct correlation between GC-poor proteomes and a lower dependence on GroEL. The authors conclude by proposing that a decrease in eubacterial GC content may have been selected in organisms facing proteostasis problems. A way to test the overall results would be through in vitro evolution experiments aimed at testing whether adaptation to low GC content provide folding advantage.The main strengths of this paper is that it addresses an interesting and timely question, finds a novel solution based on a carefully selected set of rules, and provides a clear answer. As such this article represents an excellent and elegant bioinformatics genome-wide study which will almost certainly influence our thinking about protein aggregation and evolution. Some of the weaknesses are the not always easy readability of the text which establishes unclear logical links between concepts.Another possible criticism could be that, as any in silico study, it makes strong assumptions on the sequence features that lead to aggregation and strongly relies on the quality of the classifiers used. Even though the developed classifiers seem to be more robust than previous such parameters, they remain only overall indications which can only allow statistical considerations. It could of course be argued that this is good enough to reach meaningful conclusions in this specific case.", "responses": [] } ]	1	https://f1000research.com/articles/3-137
https://f1000research.com/articles/3-136/v1	26 Jun 14	{ "type": "Research Article", "title": "Histamine and acetylcholine receptor involvement in sensorimotor gating: an autoradiography study", "authors": [ "Jerry Skefos", "Melissa Ghulam", "Arjun Mahendra", "Gaurang Patel", "Jose Larrauri", "Ehsan Kholdebarin", "Edward Levin", "Margaret Bauman", "Melissa Ghulam", "Arjun Mahendra", "Gaurang Patel", "Jose Larrauri", "Ehsan Kholdebarin", "Edward Levin", "Margaret Bauman" ], "abstract": "Sensory gating is a way by which the brain manages sensory information flow. For optimal allocation of neural resources, it is important to be able to screen out (or “gate”) irrelevant sensory information when another stimulus is being processed. Sensorimotor gating more generally refers to the overall process of modulation of the motor responses to sensory stimuli. Impaired sensorimotor gating is seen in a variety of neurobehavioral disorders including schizophrenia, autism and sensory processing disorder. The degree of sensorimotor gating can be studied behaviorally by indexing prepulse inhibition (PPI). PPI reflects the degree of suppression of a startle response to an intense sensory stimulus when it is preceded by a more modest sensory stimulus. The neural circuitry underlying PPI has been shown to include dopaminergic and cholinergic systems. We previously found that histaminergic H1 receptors also play important roles in sensorimotor gating: the acute administration of the histamine H1 antagonist, pyrilamine, significantly reverses the PPI impairment caused by the NMDA glutamate antagonist, dizocilpine (MK-801). The current study was conducted to determine the anatomic bases for histaminergic and cholinergic regulation of the effect of NMDA antagonism on PPI. Using autoradiography, we found that pyrilamine treatment decreased H1 receptor binding in the anterior cingulate, which correlated with PPI improvement. Furthermore, we found that pyrilamine treatment resulted in increased α7-nicotinic acetylcholine receptor binding in the insular cortex, which also correlated with PPI improvement. These findings shed light on the interaction between histamine and acetylcholine signaling in a distributed network of PPI modulation.", "keywords": [ "prepulse inhibition", "histaminergic H1 receptors", "nicotinic α7 receptors", "dizocilpine", "pyrilamine", "sensorimotor gating" ], "content": "Introduction\n\nSudden intense sensory stimuli can evoke a rapid muscular reaction called the startle response. The startle response can be inhibited by a milder sensory stimulus (prepulse) when presented shortly before the startling stimulus (pulse), a phenomenon termed prepulse inhibition (PPI) (Hoffman & Searle, 1965). PPI can involve any combination of visual, tactile, or auditory stimuli. PPI is believed to entail sensory gating mechanisms as well as direct motor reflex inhibition. An expansive network of brain regions underlies startle, PPI, and modulation of PPI. This network includes nuclei within the pontine reticular system, the superior and inferior colliculi, substantia nigra, basolateral amygdala, hippocampus, thalamus, prefrontal cortex, ventral pallidum, striatum, and ventral tegmental area (Swerdlow et al., 2001). In humans, PPI is believed to be related to an individual’s ability to filter incoming sensory information as well as inhibit resultant behavior (Braff et al., 2001; Rabin et al., 2009), and has been shown to be impaired in a number of neuropsychological disorders including schizophrenia and autism (Braff et al., 2001). PPI is a useful experimental technique used to investigate sensorimotor gating mechanisms and test therapeutic treatments in animal models of neurologic disease as well as in clinical populations, most extensively in patients with schizophrenia (Swerdlow et al., 2008).\n\nThe aim of this study was to investigate an established rat model of decreased PPI induced by administration of the NMDA antagonist, dizocilpine (Mansbach & Geyer, 1989), and the reversal of this PPI impairment by the histaminergic H1-antagonist, pyrilamine. H1-antagonism is a potential mechanism of the therapeutic effects of the atypical antipsychotic, clozapine, which improves PPI following dizocilpine administration in rats as well as in patients with schizophrenia (Kumari & Sharma, 2002; Levin et al., 2007; Roegge et al., 2007). In the present study we show that chronic pyrilamine administration prevents the PPI impairment induced by chronic dizocilpine administration, an effect that is correlated with a reduction in ligand-binding potential of H1 receptors in the anterior cingulate and an increase in nicotinic receptor α7 subunit binding in the insular cortex. In light of the functional anatomical connectivity of the anterior cingulate and insular cortex, both of which interact extensively with the core PPI network, our findings support the inclusion of both cortical areas in an expanded network capable of regulating sensorimotor gating.\n\n\nMethods\n\nThirty six adult female Sprague-Dawley rats (9 per treatment condition) were used for the studies (obtained from Taconic Labs, Germantown, NY). All protocols for the study were approved by the Institutional Animal Care and Use Committee (IACUC) of Duke University. The rats were housed in groups of three in a temperature-controlled vivarium on a 12 hr/12 hr reverse light-dark cycle with ad-lib access to food and water. They were tested during the dark phase.\n\nAcoustic startle reflex amplitude was measured and prepulse inhibition levels were calculated using the Med Associates Startle Reflex System (St. Albans, VT, USA). The equipment included response platforms that were placed in sound attenuating chambers. Each platform was calibrated with a spinner-type calibrator (Med Associates Startle Calibrator). A speaker was placed within the chamber midway along the long axis of the platform. The sound intensity of the speaker in each chamber was calibrated before test sessions (Digital Sound Level Meter, Extech Instruments). Plexiglas cylinders large enough to allow animals to turn around (7.5 cm diameter), were mounted on the platforms. The background white noise was a constant 65 dB.\n\nThe test session was conducted in 3 blocks. After the animals were placed in the chambers, there was a 5 min acclimation period before testing began. Block 1 consisted of 6 pulse-only trials, in which a 20 msec, 110 dB white noise stimulus (pulse) was presented. Block 2 had a total of 48 trials: 12 pulse-only trials and 36 prepulse-pulse trials. Within the prepulse-pulse trials, prepulses consisted of 20 msec pure tone noises of one of 3 possible intensities: 68, 71, or 77 dB. The trials were presented in random order with the inter-trial duration ranging from 10–20 seconds. Block 3 had an additional 5 pulse-only trials. Each stimulus had a 2 msec rise/fall time. The null period was 100 msec and the prepulse-to-pulse interval was 100 msec onset to onset. The entire test period lasted approximately 34 mins.\n\nFollowing testing, the data were analyzed to determine percent PPI for each animal. The mean percent PPI values reported were calculated in the following way:\n\n\n\nThe data from each trial were averaged for each animal, separately for each prepulse intensity, to obtain this measure of percentage inhibition of the acoustic startle responses by the prepulses.\n\nDizocilpine (MK-801, Sigma-Aldrich, St. Louis, MO) (0.15 mg/kg/day) was administered via 2ML4 osmotic minipumps for 4 weeks. The Alzet osmotic minipumps (Durect, Cupertino, CA) were implanted subcutaneously in between the scapulae on the back, in a sterile surgical procedure following manufacturer guidelines. This was covered under our approved IACUC protocol. Control animals were administered vehicle consisting of sterile 0.9% saline (Hospira, Lake Forest, IL) during this period, and pyrilamine (Sigma-Aldrich, St. Louis, MO) (50 mg/kg/day) was similarly administered with (n = 9) or without (n = 9) dizocilpine. Every week, the rats were tested for PPI, and before the final testing session in week 4, pharmacological treatments were withdrawn for one day.\n\nFollowing behavioral analyses, the rats were anesthetized with Euthasol (100 mg/kg, Virbac, Inc., Fort Worth, TX, USA) and then sacrificed by exsanguination. Brains were immediately removed and stored at -80°C. To prepare histological sections, each brain was sliced coronally at 20 µm thickness using a cryostat (Bright Instruments, Cambridgeshire, UK) and sections were mounted on gelatin-subbed slides (precleaned Superfrost Fisherbrand slides treated with 0.5% gelatin and 0.05% chromium potassium sulphate). All radioligands were purchased from American Radiolabeled Chemicals (Saint Louis, MO). Specific binding of the α7 subunit of the acetylcholine receptor was determined using [125I]α-bungarotoxin (5 nM, 147 Ci/mmol), and specific binding of the histamine H1 receptor was determined using [3H]pyrilamine (5 nM, 20 Ci/mmol). Radioligands were diluted in a chilled buffer containing 50 nM Tris-HCl and 1 mg/ml BSA, pH 7.4 and the entire radioligand binding protocol was performed on ice to limit nonspecific binding. Slides were first pre-incubated in buffer for 1 hr, after which they were incubated with the appropriate radioligand solution for 1 hr. Non-specific binding was assessed by co-incubation with l-nicotine (4 µM, Acros Organics, NJ) or nonradioactive pyrilamine (2 µM, MP Biomedicals, CA) for the α7 subunit and H1-receptor, respectively. Following incubation in radioligand solution, the slides were washed for 15 min in three changes of chilled buffer, and then subsequently dried with blown air at room temperature. Sections were then exposed to film (Kodak, Hyperfilm) for 22 days to visualize α7 subunit labeling and approximately 6 weeks to visualize pyrilamine labeling. Films were scanned (Epson, Perfection V750 Pro) and analyzed densitometrically using Image-J (Schneider et al., 2012). Each film was exposed to the appropriate set of radioactive standards, and densitometric data were converted to fmol/mg tissue equivalent.\n\nRadioligand binding was quantified within a distributed network of brain regions (Figure 1) previously implicated in regulating PPI of startle (Swerdlow et al., 2001). The regions we investigated within this network included the inferior colliculus, amygdala, and hippocampus. Within the amygdaloid complex, we sampled within specific groups of nuclei: the basolateral group, the cortical group, and the centromedial group (Sah et al., 2003), and within the hippocampal complex, we sampled CA1-3 of the hippocampus proper, as well as the dentate gyrus. Additionally, we qualified radioligand binding in the anterior cingulate and insular cortex, both of which are known to interact with the PPI network and have been preliminarily implicated in fMRI investigations of PPI in humans (Campbell et al., 2007; Kumari et al., 2007). A stereotaxic atlas (Paxinos, 2005) was used to confirm that each brain region was sampled consistently, and data from each brain hemisphere were averaged prior to statistical analysis.\n\nEach of the regions investigated with autoradiography are shown in light grey and are labeled. The colored highlighted portions of each region demonstrate the precise location within which autoradiographic analyses were performed in this study. This figure was generated from images acquired in a 3-D stereotaxic rat brain atlas software (Brain Navigator release 2.0, 2009).\n\nAnalysis of Variance (ANOVA) tests were used to assess differences between group means on the basis of pharmacologic treatment. Post-hoc comparisons were performed using Tukey’s HSD tests. Furthermore, to assess the overall effects of pyrilamine treatment, rat data were clustered into two groups (with and without pyrilamine treatment) and independent samples t-tests were performed. The appropriate t-test was chosen based on the results of Levene’s test of equality of variances between each group. For brain regions that demonstrated significant differences in radioligand binding between groups, Pearson correlations were performed to test the association between radioligand binding and PPI. All statistical tests were two-sided, with an alpha level of 0.05. Confidence intervals were set at 95% for all comparisons. Statistical computations were performed with SPSS Version 19 (IBM, Armonk, NY, USA).\n\n\nResults\n\nAlthough prepulse intensity did not significantly affect inhibition of the startle response, we did observe a trend towards increased PPI with increased prepulse intensity (Figure 2, Dataset 1). For trials using the 77 dB prepulse, the highest intensity employed, ANOVA demonstrated a significant effect of pharmacologic treatment (p = 0.026), with pyrilamine administration significantly attenuating the dizocilpine-induced PPI impairment (p = 0.021).\n\nAt each prepulse decibel (dB) level employed, the general pattern of pharmacologic manipulation of PPI was maintained: pyrilamine treatment increased PPI and attenuated dizocilpine-induced PPI impairment. Furthermore, in agreement with prior reports, increased prepulse intensity was associated with stronger prepulse inhibition. Error bars represent standard error of the mean.\n\nWithin the anterior cingulate, pyrilamine treatment induced a 37 ± 19% decrease in H1-receptor binding that was nearly significant (p = 0.058) (Figure 3, Dataset 2). This change correlated with PPI such that decreased H1-receptor binding was associated with increased mean PPI at all three prepulse intensities (Figure 4, R2 = -0.15, p = 0.05). Note as demonstrated in Figure 4, Figure 6, and Figure 7 that some rats (irrespective of pharmacologic treatment) did not demonstrate the PPI phenomenon: this is a common occurrence in PPI studies and reflects inter-individual differences.\n\nAn overall effect of pyrilamine treatment was a reduction in H1-receptor binding in the anterior cingulate (AC). This was a nearly significant trend that will require further investigation to confirm (p = 0.058).\n\nThere was a negative correlation between H1 receptor binding in the anterior cingulate (AC) and PPI, such that lower H1 binding in the AC was associated with higher PPI. Data are displayed for the 77 dB prepulse trials for which PPI was greatest.\n\nWithin the insular cortex, pyrilamine treatment induced a 9 ± 3% increase in α7 acetylcholine (ACh)-receptor binding (p = 0.002) (Figure 5, Dataset 2). Increased α7 ACh-receptor binding correlated with increased mean PPI at all three prepulse intensities (Figure 6, R2 = 0.24, p = 0.004).\n\nAn overall effect of pyrilamine treatment was an increase in α7-nicotinic receptor binding in the insular cortex (p = 0.002).\n\nWe observed a positive correlation between α7-nicotinic receptor binding in the insular cortex and PPI, such that greater α7 binding in the insula was associated with higher PPI. Data are displayed for the 77 dB prepulse trials for which PPI was greatest.\n\nThe results from this study imply that pyrilamine modulates PPI of the startle response through diverse mechanisms that are regionally discrete. In the anterior cingulate, α7 ACh-receptor binding was positively correlated with mean PPI (Figure 7, R2 = 0.14, p = 0.027); however, α7 ACh-receptor binding in this region was not significantly affected by pyrilamine treatment (p = 0.507, Dataset 2). Likewise, pyrilamine treatment did not affect H1-receptor binding in the insular cortex (p = 0.237), and H1-receptor binding in the insula was not associated with mean PPI (p = 0.702). In the other brain regions investigated (inferior colliculus, amygdala, and dorsal hippocampus), neither H1-receptor binding nor α7 ACh-receptor binding was associated with PPI levels.\n\nAlthough pyrilamine treatment did not affect α7 receptor binding in the anterior cingulate (AC), we did observe a positive correlation between α7 receptor binding in this region and PPI. This indicates that α7 receptors in the AC may contribute to PPI independently of H1 receptor-mediated signaling.\n\n\n\n\nDiscussion\n\nSensorimotor gating impairment has been associated with a wide variety of neurological and psychiatric disorders. It has been observed in patients with sensory processing disorder (Davies et al., 2009), Parkinson’s disease (Nakashima et al., 1993), schizophrenia (Braff et al., 2001), non-epileptic seizures (Pouretemad et al., 1998), Tourette’s syndrome and ADHD (Castellanos et al., 1996), nocturnal enuresis (Ornitz et al., 1992), blepharospasm (Gómez-Wong et al., 1998), obsessive-compulsive disorder (Swerdlow et al., 1993; Hoenig et al., 2005), panic disorder (Ludewig et al., 2002), bipolar disorder (Perry et al., 2001; Rich et al., 2005), Huntington’s disease (Swerdlow et al., 1995), Fragile X syndrome (Hessl et al., 2009), and autism (Perry et al., 2007). In humans, sensory gating measures such as PPI have been found to correlate with a growing list of behavioral symptoms, including neuroticism, disinhibition on the go/no-go task, premonitory urges in Tourette’s syndrome, high trait anxiety in panic disorder, restricted/repetitive behavior in autism, and in schizophrenia: positive and negative symptoms, semantic priming abnormalities, perseveration, information-processing deficits, thought disorder, distractibility, and violent behavior (Perry & Braff, 1994; Karper et al., 1996; Vinogradov et al., 1996; Braff et al., 1999; Corr et al., 2002; Ludewig et al., 2002; Perry et al., 2007; Rabin et al., 2009; Yadon et al., 2009). It is important to investigate the neural bases that underlie dysfunctions of sensory gating in order to develop effective therapeutic treatments.\n\nThe present study employed a PPI paradigm in which an auditory startle stimulus was immediately preceded by a milder auditory prepulse expected to induce significant gating of the startle response. Chronic administration of the NMDA antagonist, dizocilpine, caused a considerable deficit in sensorimotor gating, demonstrated by the low levels of PPI of the startle response in dizocilpine-treated rats (mean PPI of 5.6%). This deficit was entirely reversed by co-administration of the selective histaminergic H1-receptor antagonist, pyrilamine (Figure 2).\n\nDizocilpine disruption of PPI has been previously used to model sensorimotor gating impairments (Mansbach & Geyer, 1989; Geyer et al., 2001). The predictive validity of this model is supported by the fact that the atypical antipsychotic, clozapine, reverses PPI impairment in this model (Levin et al., 2007; Lim et al., 2012) as it does similarly in humans (Nagamoto et al., 1996; Adler et al., 1998). Here we show that chronic pyrilamine treatment mimics the effect of clozapine, reversing PPI impairment in the dizocilpine-treated rat. As clozapine has been shown to saturate the brain’s H1-receptors at therapeutic concentrations (Humbert-Claude et al., 2012), this may provide an important mechanism of clozapine’s therapeutic action.\n\nIn the mammalian brain, histaminergic fibers originating in the tuberomammillary nucleus of the posterior hypothalamus regulate the response to noxious stimuli (Itoh et al., 1989) and infection (Saper et al., 2012), and also regulate the excitability of arousal circuits (Tasaka et al., 1989). One mechanism of histamine’s effects on arousal is through regulation of cholinergic transmission in the nucleus basalis of Myenert in the basal forebrain (Bacciottini et al., 2001; Dringenberg & Kuo, 2003). Histamine’s cholinergic effects have been shown to differ regionally in the brain: in the frontoparietal cortex, local histamine administration inhibited acetylcholine release by 50% (Blandina et al., 1996), whereas histamine application to the nucleus basalis has been shown to double cholinergic output to the cortex in rats (Cecchi et al., 1998). The complicated interactions between histaminergic signaling and cholinergic tone throughout the brain awaits further elucidation; however, it has been shown that tuberomammillary lesions or systemic administration of the H1 antagonists, chlorpheniramine and pyrilamine, are similarly capable of significantly increasing cortical acetylcholine dose-dependently (Dringenberg et al., 1998). Further, members of our group have recently shown that pyrilamine treatment reduces nicotine self-administration in rats (Levin et al., 2011). So it is believed that systemic H1 antagonists modify behavior in part by increasing cortical cholinergic transmission.\n\nThe role of histaminergic signaling in PPI has not been extensively studied, but of the disorders investigated to date that are associated with sensorimotor gating abnormalities, all have displayed increased histaminergic neurotransmission. These include: Parkinson’s disease (Rinne et al., 2002), Tourette’s syndrome (Fernandez et al., 2012), bipolar disorder (Jin et al., 2009), Huntington’s disease (van Wamelen et al., 2011), and schizophrenia (Ito, 2004; Arrang, 2007). In Alzheimer’s dementia, interestingly, researchers found significantly decreased histamine levels in the frontal and temporal cortices (Mazurkiewicz-Kwilecki & Nsonwah, 1989), and normal PPI (Hejl et al., 2004).\n\nPrior work by members of our group demonstrated the ability of a single dose of pyrilamine to attenuate PPI impairment in rats acutely administered dizocilpine or amphetamine (Roegge et al., 2007; Larrauri & Levin, 2010). The current study was designed to further explore this phenomenon pharmacologically in rats by chronically administering dizocilpine and/or pyrilamine, a scenario more analogous to long-term therapeutic enhancement of sensorimotor gating in humans. Our aim was to determine which components of the distributed PPI network were impacted by pyrilamine treatment to improve PPI.\n\nPyrilamine treatment resulted in decreased H1-receptor binding in the anterior cingulate, which was correlated with PPI improvement (Figure 3 and Figure 4). Functional involvement of the anterior cingulate in PPI has been previously demonstrated through lesion as well as fMRI and PET studies (Hazlett et al., 1998; Yee, 2000; Goldman et al., 2006; Campbell et al., 2007; Neuner et al., 2010), and this region is known to express a high concentration of H1-receptors (Tagawa et al., 2001). As a site of limbic and cortical integration, the anterior cingulate modulates conditioned fear responses and arousal (Hamner et al., 1999). The anterior cingulate has been shown to send efferent projections to the amygdala (Wang et al., 2009), a wide distribution of thalamic nuclei (Fujii, 1983), substantia nigra (Beckstead, 1979), nucleus accumbens (Sesack et al., 1989), globus pallidus (Beckstead, 1979), and superior colliculus (Sesack et al., 1989), all regions implicated in regulating PPI of the startle response (Yamada et al., 1998; Fendt et al., 2001; Hazlett et al., 2001; Takahashi et al., 2007; Forcelli et al., 2012). Efferents of the anterior cingulate have even been traced to the giant neurons of the caudal pontine reticular nucleus, which display prepulse-inhibited membrane potential and are believed to be an integral component of the acoustic startle response (Sesack et al., 1989; Lingenhöhl & Friauf, 1994). Similarly, anterior cingulate efferents were traced to the adjacent pedunculopontine tegmental nucleus (Sesack et al., 1989), which is an established component of the core brainstem PPI circuitry that directly modulates the pontine reticular giant neurons through cholinergic innervation (Fendt et al., 2001). We have shown that α7 ACh-receptor binding in the anterior cingulate is positively correlated with PPI (Figure 7), so it is possible that cholinergic signaling in this region plays a role in PPI. Indeed, nicotine has been shown to increase activity in the anterior cingulate while improving PPI (Postma et al., 2006). However, we found no effect of pyrilamine on the α7 ACh-receptor binding in the anterior cingulate, implicating an alternate mechanism of pyrilamine enhancement of PPI in this brain region.\n\nContrary to the anterior cingulate, the insular cortex displayed a significant increase in α7 ACh-receptor binding in pyrilamine-treated rats that was positively correlated with PPI (Figure 5 and Figure 6). However, there was no change in H1-receptor binding in the insula of rats that were treated with pyrilamine. Because H1 antagonists have been shown to be ineffective in regulating insular cortical excitability when directly applied to this region (Takei et al., 2012), pyrilamine treatment may modulate insular acetylcholine signaling indirectly. The insular cortex contains a high concentration of nicotinic receptors, and has shown increased activation following nicotine administration in patients with schizophrenia whose PPI was improved by nicotine (Postma et al., 2006). Agonism of α7 ACh-receptors has been demonstrated to improve sensory gating in patients with schizophrenia (Martin & Freedman, 2007), and has proven beneficial in animal models of not only gating impairments (Cilia et al., 2005; Hajós et al., 2005; Thomsen et al., 2009) but also positive and negative symptoms of schizophrenia (Hauser et al., 2009). Therefore, it is likely that pyrilamine-induced increases in insular α7 ACh-receptor expression contribute to the observed improvement in PPI.\n\nDue to its anatomic connections, the insular cortex is believed to integrate the processing of autonomic responses with that of ongoing behavioral plans and emotional states (Allen et al., 1991; Gu et al., 2013). The insula is anatomically poised to influence top-down processing of sensorimotor gating, with both direct as well as indirect connections to the pontine reticular startle circuit (Wiesendanger & Wiesendanger, 1982). The insula send largely reciprocal projections to the amygdala and mediodorsal nucleus of the thalamus (Shi & Cassell, 1998), substantia nigra, raphe nucleus, ventral pallidum, and ventral striatum (Reep & Winans, 1982), all of which have been associated with regulation of PPI (Young et al., 1995; Fendt et al., 2001; Adams et al., 2008; Baldan et al., 2011; Forcelli et al., 2012). The insula have furthermore been shown to be involved in distinguishing successive stimuli presented with a short interstimulus time interval as is required to elicit PPI (Kosillo & Smith, 2010).\n\nAlthough the anterior cingulate and insular cortex have been shown to be highly interconnected functionally (Di Martino et al., 2009; Medford & Critchley, 2010; Cauda et al., 2011), there was no correlation between the changes we report in pyrilamine binding in the anterior cingulate and α-bungarotoxin binding in the insular cortex. The anterior cingulate and insular cortex both send projections to regions that are known to directly modulate the startle response through interaction with the pontine reticular nuclei responsible for this reflex. It is therefore possible that systemic pyrilamine treatment modified the activity of the anterior cingulate and insular cortex independently to improve PPI, although further research is necessary to understand the interaction between anterior cingulate and insular cortical networks in modulating startle inhibition. This study demonstrates the wide distribution of networks capable of influencing PPI of startle, supporting the importance of the PPI measure as a tool to analyze interactions between multiple hierarchies of neuronal processing in disparate brain regions, particularly in pathological states. It will be important to further investigate effects of pharmacological treatments on PPI in humans as well as animal models of disease to elucidate the neuronal machinery underlying sensory filtering and behavioral inhibition. The involvement of H1 receptors in PPI has not only been demonstrated in pharmacologically induced PPI disruption models, but also in models of developmental disturbance that show PPI impairment. In a mouse model of isolation rearing, pyrilamine treatment or H1 receptor knockout were capable of preventing PPI impairment (Dai et al., 2005) similarly to clozapine (Möller et al., 2011). Along these lines, further research into the histaminergic regulation of PPI is warranted, as H1-antagonists have demonstrated a low incidence of side effects in clinical trials (Pearlman et al., 1997; Lankford et al., 2012). A recent initial clinical study has shown that the antihistamine meclizine significantly improves PPI in people who have less than typical PPI and exaggerated startle response (Larrauri et al., 2014). H1 antagonists may therefore prove useful in reversing sensory gating dysfunctions. Further evaluation should determine the possible efficacy and side effects of this novel line of treatment.\n\n\nData availability\n\nfigshare: Prepulse inhibition of the startle response and radioligand binding assays, doi: 10.6084/m9.figshare.1060215 (Skefos et al., 2014)", "appendix": "Author contributions\n\n\n\nEDL and MLB conceived the study. EDL, MLB and JS designed the experiments. JL and EK carried out the animal surgery, behavioral testing, and dissection. JS, MG, AM, and GP carried out the tissue preparation, radioligand binding protocols, and quantification. JS carried out the data analyses and prepared the first draft of the manuscript. EDL, MLB, JL, and JS revised the manuscript. All authors provided input in the design of this study and the reporting of the findings.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nFunds were assigned to M.L. Bauman and E.D. Levin by the Wallace Research Foundation and The Autism Research Foundation to perform the experiments here presented.\n\n\nReferences\n\nAdams W, Kusljic S, van den Buuse M: Serotonin depletion in the dorsal and ventral hippocampus: effects on locomotor hyperactivity, prepulse inhibition and learning and memory. Neuropharmacology. 2008; 55(6): 1048–1055. PubMed Abstract \| Publisher Full Text\n\nAdler LE, Olincy A, Waldo M, et al.: Schizophrenia, sensory gating, and nicotinic receptors. Schizophr Bull. 1998; 24(2): 189–202. PubMed Abstract\n\nAllen GV, Saper CB, Hurley KM, et al.: Organization of visceral and limbic connections in the insular cortex of the rat. J Comp Neurol. 1991; 311(1): 1–16. PubMed Abstract \| Publisher Full Text\n\nArrang JM: Histamine and Schizophrenia. Int Rev Neurobiol. Academic Press. 2007; 78: 247–287. PubMed Abstract \| Publisher Full Text\n\nBacciottini L, Passani MB, Mannaioni PF, et al.: Interactions between histaminergic and cholinergic systems in learning and memory. Behav Brain Res. 2001; 124(2): 183–194. PubMed Abstract \| Publisher Full Text\n\nBaldan Ramsey LC, Xu M, Wood N, et al.: Lesions of the dorsomedial striatum disrupt prepulse inhibition. Neuroscience. 2011; 180: 222–228. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeckstead RM: An autoradiographic examination of corticocortical and subcortical projections of the mediodorsal-projection (prefrontal) cortex in the rat. J Comp Neurol. 1979; 184(1): 43–62. PubMed Abstract \| Publisher Full Text\n\nBlandina P, Giorgetti M, Bartolini L, et al.: Inhibition of cortical acetylcholine release and cognitive performance by histamine H3 receptor activation in rats. Br J Pharmacol. 1996; 119(8): 1656–1664. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBraff DL, Geyer MA, Swerdlow NR: Human studies of prepulse inhibition of startle: normal subjects, patient groups, and pharmacological studies. Psychopharmacology. 2001; 156(2–3): 234–258. PubMed Abstract \| Publisher Full Text\n\nBraff DL, Swerdlow NR, Geyer MA: Symptom correlates of prepulse inhibition deficits in male schizophrenic patients. Am J Psychiatry. 1999; 156(4): 596–602. PubMed Abstract\n\nCampbell LE, Hughes M, Budd TW, et al.: Primary and secondary neural networks of auditory prepulse inhibition: a functional magnetic resonance imaging study of sensorimotor gating of the human acoustic startle response. Eur J Neurosci. 2007; 26(8): 2327–2333. PubMed Abstract \| Publisher Full Text\n\nCastellanos FX, Fine EJ, Kaysen D, et al.: Sensorimotor gating in boys with Tourette’s syndrome and ADHD: preliminary results. Biol Psychiatry. 1996; 39(1): 33–41. PubMed Abstract \| Publisher Full Text\n\nCauda F, D'Agata F, Sacco K, et al.: Functional connectivity of the insula in the resting brain. NeuroImage. 2011; 55(1): 8–23. PubMed Abstract \| Publisher Full Text\n\nCecchi M, Giorgetti M, Bacciottini L, et al.: Increase of acetylcholine release from cortex of freely moving rats by administration of histamine into the nucleus basalis magnocellularis. Inflamm Res. 1998; 47(Suppl 1): S32–S33. PubMed Abstract \| Publisher Full Text\n\nChikama M, McFarland NR, Amaral DG, et al.: Insular cortical projections to functional regions of the striatum correlate with cortical cytoarchitectonic organization in the primate. J Neurosci. 1997; 17(24): 9686–9705. PubMed Abstract\n\nCilia J, Cluderay JE, Robbins MJ, et al.: Reversal of isolation-rearing-induced PPI deficits by an alpha7 nicotinic receptor agonist. Psychopharmacology (Berl). 2005; 182(2): 214–219. PubMed Abstract \| Publisher Full Text\n\nCorr PJ, Tynan A, Kumari V: Personality correlates of prepulse inhibition of the startle reflex at three lead intervals. J Psychophysiol. 2002; 16(2): 82–91. Publisher Full Text\n\nDai H, Okuda T, Sakurai E, et al.: Blockage of histamine H1 receptor attenuates social isolation-induced disruption of prepulse inhibition: a study in H1 receptor gene knockout mice. Psychopharmacology (Berl). 2005; 183(3): 285–293. PubMed Abstract \| Publisher Full Text\n\nDavies PL, Chang WP, Gavin WJ: Maturation of sensory gating performance in children with and without sensory processing disorders. Int J Psychophysiol. 2009; 72(2): 187–197. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDevinsky O, Morrell MJ, Vogt BA: Contributions of anterior cingulate cortex to behaviour. Brain. 1995; 118(1): 279–306. PubMed Abstract \| Publisher Full Text\n\nDi Martino A, Shehzad Z, Kelly C, et al.: Relationship between cingulo-insular functional connectivity and autistic traits in neurotypical adults. Am J Psychiatry. 2009; 166(8): 891–899. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDringenberg HC, De Souza-Silva MA, Rossmüller J, et al.: Histamine H1 receptor antagonists produce increases in extracellular acetylcholine in rat frontal cortex and hippocampus. J Neurochem. 1998; 70(4): 1750–1758. PubMed Abstract \| Publisher Full Text\n\nDringenberg HC, Kuo MC: Histaminergic facilitation of electrocorticographic activation: role of basal forebrain, thalamus, and neocortex. Eur J Neurosci. 2003; 18(8): 2285–2291. PubMed Abstract \| Publisher Full Text\n\nDu Y, Wu X, Li L: Differentially organized top-down modulation of prepulse inhibition of startle. J Neurosci. 2011; 31(38): 13644–13653. PubMed Abstract \| Publisher Full Text\n\nFendt M, Li L, Yeomans JS: Brain stem circuits mediating prepulse inhibition of the startle reflex. Psychopharmacology (Berl). 2001; 156(2–3): 216–224. PubMed Abstract \| Publisher Full Text\n\nFernandez TV, Sanders SJ, Yurkiewicz IR, et al.: Rare copy number variants in tourette syndrome disrupt genes in histaminergic pathways and overlap with autism. Biol Psychiatry. 2012; 71(5): 392–402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForcelli PA, West EA, Murnen AT, et al.: Ventral pallidum mediates amygdala-evoked deficits in prepulse inhibition. Behav Neurosci. 2012; 126(2): 290–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFujii M: Fiber connections between the thalamic posterior lateral nucleus and the cingulate gyrus in the cat. Neurosci Lett. 1983; 39(2): 137–142. PubMed Abstract \| Publisher Full Text\n\nGeyer MA, Krebs-Thomson K, Braff DL, et al.: Pharmacological studies of prepulse inhibition models of sensorimotor gating deficits in schizophrenia: a decade in review. Psychopharmacology (Berl). 2001; 156(2–3): 117–154. PubMed Abstract \| Publisher Full Text\n\nGoldman MB, Heidinger L, Kulkarni K, et al.: Changes in the amplitude and timing of the hemodynamic response associated with prepulse inhibition of acoustic startle. NeuroImage. 2006; 32(3): 1375–1384. PubMed Abstract \| Publisher Full Text\n\nGómez-Wong E, Martí MJ, Tolosa E, et al.: Sensory modulation of the blink reflex in patients with blepharospasm. Arch Neurol. 1998; 55(9): 1233–1237. PubMed Abstract\n\nGu X, Liu X, Van Dam NT, et al.: Cognition-emotion Integration in the anterior insular cortex. Cereb Cortex. 2013; 23(1): 20–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHajós M, Hurst RS, Hoffmann WE, et al.: The selective alpha7 nicotinic acetylcholine receptor agonist PNU-282987 [N-[(3R)-1-Azabicyclo[2.2.2]oct-3-yl]-4-chlorobenzamide hydrochloride] enhances GABAergic synaptic activity in brain slices and restores auditory gating deficits in anesthetized rats. J Pharmacol Exp Ther. 2005; 312(3): 1213–1222. PubMed Abstract \| Publisher Full Text\n\nHamner MB, Lorberbaum JP, George MS: Potential role of the anterior cingulate cortex in PTSD: review and hypothesis. Depress Anxiety. 1999; 9(1): 1–14. PubMed Abstract \| Publisher Full Text\n\nHauser TA, Kucinski A, Jordan KG, et al.: TC-5619: an alpha7 neuronal nicotinic receptor-selective agonist that demonstrates efficacy in animal models of the positive and negative symptoms and cognitive dysfunction of schizophrenia. Biochem Pharmacol. 2009; 78(7): 803–812. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHazlett EA, Buchsbaum MS, Haznedar MM, et al.: Prefrontal cortex glucose metabolism and startle eyeblink modification abnormalities in unmedicated schizophrenia patients. Psychophysiology. 1998; 35(2): 186–198. PubMed Abstract \| Publisher Full Text\n\nHazlett EA, Buchsbaum MS, Tang CY, et al.: Thalamic activation during an attention-to-prepulse startle modification paradigm: a functional MRI study. Biol Psychiatry. 2001; 50(4): 281–291. PubMed Abstract \| Publisher Full Text\n\nHejl AM, Glenthøj B, Mackeprang T, et al.: Prepulse inhibition in patients with Alzheimer’s disease. Neurobiol Aging. 2004; 25(8): 1045–1050. PubMed Abstract \| Publisher Full Text\n\nHessl D, Berry-Kravis E, Cordeiro L, et al.: Prepulse inhibition in fragile X syndrome: feasibility, reliability, and implications for treatment. Am J Med Genet B Neuropsychiatr Genet. 2009; 150B(4): 545–553. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoenig K, Hochrein A, Quednow BB, et al.: Impaired prepulse inhibition of acoustic startle in obsessive-compulsive disorder. Biol Psychiatry. 2005; 57(10): 1153–1158. PubMed Abstract \| Publisher Full Text\n\nHoffman HS, Searle JL: Acoustic variables in the modification of the startle reaction in the rat. J Comp Physiol Psychol. 1965; 60: 53–58. PubMed Abstract \| Publisher Full Text\n\nHumbert-Claude M, Davenas E, Gbahou F, et al.: Involvement of histamine receptors in the atypical antipsychotic profile of clozapine: a reassessment in vitro and in vivo. Psychopharmacology (Berl). 2012; 220(1): 225–241. PubMed Abstract \| Publisher Full Text\n\nIto C: The role of the central histaminergic system on schizophrenia. Drug News Perspect. 2004; 17(6): 383–387. PubMed Abstract \| Publisher Full Text\n\nItoh Y, Oishi R, Nishibori M, et al.: Effects of nociceptive stimuli on brain histamine dynamics. Jpn J Pharmacol. 1989; 49(4): 449–454. PubMed Abstract \| Publisher Full Text\n\nJin CY, Anichtchik O, Panula P: Altered histamine H3 receptor radioligand binding in post-mortem brain samples from subjects with psychiatric diseases. Br J Pharmacol. 2009; 157(1): 118–129. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarper LP, Freeman GK, Grillon C, et al.: Preliminary evidence of an association between sensorimotor gating and distractibility in psychosis. J Neuropsychiatry Clin Neurosci. 1996; 8(1): 60–66. PubMed Abstract\n\nKnigge U, Kjaer A, Jørgensen H, et al.: Role of hypothalamic histaminergic neurons in mediation of ACTH and beta-endorphin responses to LPS endotoxin in vivo. Neuroendocrinology. 1994; 60(3): 243–251. PubMed Abstract \| Publisher Full Text\n\nKoliatsos VE, Martin LJ, Walker LC, et al.: Topographic, non-collateralized basal forebrain projections to amygdala, hippocampus, and anterior cingulate cortex in the rhesus monkey. Brain Res. 1988; 463(1): 133–139. PubMed Abstract \| Publisher Full Text\n\nKosillo P, Smith AT: The role of the human anterior insular cortex in time processing. Brain Struct Funct. 2010; 214(5–6): 623–628. PubMed Abstract \| Publisher Full Text\n\nKumari V, Antonova E, Geyer MA, et al.: A fMRI investigation of startle gating deficits in schizophrenia patients treated with typical or atypical antipsychotics. Int J Neuropsychopharmacol. 2007; 10(4): 463–477. PubMed Abstract \| Publisher Full Text\n\nKumari V, Sharma T: Effects of typical and atypical antipsychotics on prepulse inhibition in schizophrenia: a critical evaluation of current evidence and directions for future research. Psychopharmacology. 2002; 162(2): 97–101. PubMed Abstract \| Publisher Full Text\n\nLankford A, Rogowski R, Essink B, et al.: Efficacy and safety of doxepin 6 mg in a four-week outpatient trial of elderly adults with chronic primary insomnia. Sleep Med. 2012; 13(2): 133–138. PubMed Abstract \| Publisher Full Text\n\nLarrauri JA, Levin ED: PPI deficit induced by amphetamine is attenuated by the histamine H1 antagonist pyrilamine, but is exacerbated by the serotonin 5-HT2 antagonist ketanserin. Psychopharmacology (Berl). 2010; 212(4): 551–558. PubMed Abstract \| Publisher Full Text\n\nLarrauri JA, Kelley LD, Jenkins MR, et al.: Meclizine enhancement of sensorimotor gating in healthy male subjects with high startle responses and low prepulse inhibition. Neuropsychopharmacology. 2014; 39(3): 651–659. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLevin ED, Caldwell DP, Perraut C: Clozapine treatment reverses dizocilpine-induced deficits of pre-pulse inhibition of tactile startle response. Pharmacol Biochem Behav. 2007; 86(3): 597–605. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLevin ED, Slade S, Wells C, et al.: Histamine H(1) antagonist treatment with pyrilamine reduces nicotine self-administration in rats. Eur J Pharmacol. 2011; 650(1): 256–260. PubMed Abstract \| Publisher Full Text\n\nLim AL, Taylor DA, Malone DT: A two-hit model: behavioural investigation of the effect of combined neonatal MK-801 administration and isolation rearing in the rat. J Psychopharmacol. 2012; 26(9): 1252–1264. PubMed Abstract \| Publisher Full Text\n\nLingenhöhl K, Friauf E: Giant neurons in the rat reticular formation: a sensorimotor interface in the elementary acoustic startle circuit? J Neurosci. 1994; 14(3 Pt 1): 1176–1194. PubMed Abstract\n\nLudewig S, Ludewig K, Geyer MA, et al.: Prepulse inhibition deficits in patients with panic disorder. Depress Anxiety. 2002; 15(2): 55–60. PubMed Abstract \| Publisher Full Text\n\nMansbach RS, Geyer MA: Effects of phencyclidine and phencyclidine biologs on sensorimotor gating in the rat. Neuropsychopharmacology. 1989; 2(4): 299–308. PubMed Abstract \| Publisher Full Text\n\nMartin LF, Freedman R: Schizophrenia and the alpha7 nicotinic acetylcholine receptor. Int Rev Neurobiol. 2007; 78: 225–246. PubMed Abstract \| Publisher Full Text\n\nMazurkiewicz-Kwilecki IM, Nsonwah S: Changes in the regional brain histamine and histidine levels in postmortem brains of Alzheimer patients. Can J Physiol Pharmacol. 1989; 67(1): 75–78. PubMed Abstract \| Publisher Full Text\n\nMedford N, Critchley HD: Conjoint activity of anterior insular and anterior cingulate cortex: awareness and response. Brain Struct Funct. 2010; 214(5–6): 535–549. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMöller M, Du Preez JL, Emsley R, et al.: Isolation rearing-induced deficits in sensorimotor gating and social interaction in rats are related to cortico-striatal oxidative stress, and reversed by sub-chronic clozapine administration. Eur Neuropsychopharmacol. 2011; 21(6): 471–483. PubMed Abstract \| Publisher Full Text\n\nNagamoto HT, Adler LE, Hea RA, et al.: Gating of auditory P50 in schizophrenics: unique effects of clozapine. Biol Psychiatry. 1996; 40(3): 181–188. PubMed Abstract \| Publisher Full Text\n\nNakashima K, Shimoyama R, Yokoyama Y, et al.: Auditory effects on the electrically elicited blink reflex in patients with Parkinson's disease. Electroencephalogr Clin Neurophysiol. 1993; 89(2): 108–112. PubMed Abstract \| Publisher Full Text\n\nNeuner I, Stöcker T, Kellermann T, et al.: Electrophysiology meets fMRI: neural correlates of the startle reflex assessed by simultaneous EMG-fMRI data acquisition. Hum Brain Mapp. 2010; 31(11): 1675–1685. PubMed Abstract \| Publisher Full Text\n\nOrnitz EM, Hanna GL, de Traversay J: Prestimulation-induced startle modulation in attention-deficit hyperactivity disorder and nocturnal enuresis. Psychophysiology. 1992; 29(4): 437–451. PubMed Abstract \| Publisher Full Text\n\nPaxinos G: The rat brain in stereotaxic coordinates. Amsterdam; Boston, Elsevier Academic Press. 2005. Reference Source\n\nPearlman DS, Lumry WR, Winder JA, et al.: Once-daily cetirizine effective in the treatment of seasonal allergic rhinitis in children aged 6 to 11 years: a randomized, double-blind, placebo-controlled study. Clin Pediatr (Phila). 1997; 36(4): 209–215. PubMed Abstract \| Publisher Full Text\n\nPerry W, Braff DL: Information-processing deficits and thought disorder in schizophrenia. Am J Psychiatry. 1994; 151(3): 363–367. PubMed Abstract\n\nPerry W, Minassian A, Feifel D, et al.: Sensorimotor gating deficits in bipolar disorder patients with acute psychotic mania. Biol Psychiatry. 2001; 50(6): 418–424. PubMed Abstract \| Publisher Full Text\n\nPerry W, Minassian A, Lopez B, et al.: Sensorimotor gating deficits in adults with autism. Biol Psychiatry. 2007; 61(4): 482–486. PubMed Abstract \| Publisher Full Text\n\nPetryshen TL, Kirby A, Hammer RP Jr, et al.: Two quantitative trait loci for prepulse inhibition of startle identified on mouse chromosome 16 using chromosome substitution strains. Genetics. 2005; 171(4): 1895–1904. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPostma P, Gray JA, Sharma T, et al.: A behavioural and functional neuroimaging investigation into the effects of nicotine on sensorimotor gating in healthy subjects and persons with schizophrenia. Psychopharmacology (Berl). 2006; 184(3–4): 589–599. PubMed Abstract \| Publisher Full Text\n\nPouretemad HR, Thompson PJ, Fenwick PB: Impaired sensorimotor gating in patients with non-epileptic seizures. Epilepsy Res. 1998; 31(1): 1–12. PubMed Abstract \| Publisher Full Text\n\nRabin RA, Sacco KA, George TP: Correlation of prepulse inhibition and Wisconsin Card Sorting Test in schizophrenia and controls: effects of smoking status. Schizophr Res. 2009; 114(1–3): 91–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReep RL, Winans SS: Efferent connections of dorsal and ventral agranular insular cortex in the hamster, Mesocricetus auratus. Neuroscience. 1982; 7(11): 2609–2635. PubMed Abstract \| Publisher Full Text\n\nRich BA, Vinton D, Grillon C, et al.: An investigation of prepulse inhibition in pediatric bipolar disorder. Bipolar Disord. 2005; 7(2): 198–203. PubMed Abstract \| Publisher Full Text\n\nRinne JO, Anichtchik OV, Eriksson KS, et al.: Increased brain histamine levels in Parkinson's disease but not in multiple system atrophy. J Neurochem. 2002; 81(5): 954–960. PubMed Abstract \| Publisher Full Text\n\nRoegge CS, Perraut C, Hao X, et al.: Histamine H1 receptor involvement in prepulse inhibition and memory function: relevance for the antipsychotic actions of clozapine. Pharmacol Biochem Behav. 2007; 86(4): 686–692. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSah P, Faber ES, Lopez De Armentia M, et al.: The amygdaloid complex: anatomy and physiology. Physiol Rev. 2003; 83(3): 803–834. PubMed Abstract\n\nSaper CB: Convergence of autonomic and limbic connections in the insular cortex of the rat. J Comp Neurol. 1982; 210(2): 163–173. PubMed Abstract \| Publisher Full Text\n\nSaper CB, Romanovsky AA, Scammell TE: Neural circuitry engaged by prostaglandins during the sickness syndrome. Nat Neurosci. 2012; 15(8): 1088–1095. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012; 9(7): 671–675. PubMed Abstract \| Publisher Full Text\n\nSesack SR, Deutch AY, Roth RH, et al.: Topographical organization of the efferent projections of the medial prefrontal cortex in the rat: an anterograde tract-tracing study with Phaseolus vulgaris leucoagglutinin. J Comp Neurol. 1989; 290(2): 213–242. PubMed Abstract \| Publisher Full Text\n\nShi CJ, Cassell MD: Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. J Comp Neurol. 1998; 399(4): 440–468. PubMed Abstract \| Publisher Full Text\n\nShibata H, Naito J: Organization of anterior cingulate and frontal cortical projections to the anterior and laterodorsal thalamic nuclei in the rat. Brain Res. 2005; 1059(1): 93–103. PubMed Abstract \| Publisher Full Text\n\nSkefos J, Ghulam M, Mahendra A, et al.: Prepulse inhibition of the startle response and radioligand binding assays. figshare. 2014. Data Source\n\nSutherland Owens AN, Miguel EC, Swerdlow NR: Sensory gating scales and premonitory urges in tourette syndrome. ScientificWorldJournal. 2011; 11: 736–741. PubMed Abstract \| Publisher Full Text\n\nSwerdlow NR, Benbow CH, Zisook S, et al.: A preliminary assessment of sensorimotor gating in patients with obsessive compulsive disorder. Biol Psychiatry. 1993; 33(4): 298–301. PubMed Abstract \| Publisher Full Text\n\nSwerdlow NR, Geyer MA, Braff DL: Neural circuit regulation of prepulse inhibition of startle in the rat: current knowledge and future challenges. Psychopharmacology (Berl). 2001; 156(2–3): 194–215. PubMed Abstract \| Publisher Full Text\n\nSwerdlow NR, Paulsen J, Braff DL, et al.: Impaired prepulse inhibition of acoustic and tactile startle response in patients with Huntington’s disease. J Neurol Neurosurg Psychiatry. 1995; 58(2): 192–200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSwerdlow NR, Weber M, Qu Y, et al.: Realistic expectations of prepulse inhibition in translational models for schizophrenia research. Psychopharmacology (Berl). 2008; 199(3): 331–388. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTagawa M, Kano M, Okamura N, et al.: Neuroimaging of histamine H1-receptor occupancy in human brain by positron emission tomography (PET): A comparative study of ebastine, a second-generation antihistamine, and (+)-chlorpheniramine, a classical antihistamine. Br J Clin Pharmacol. 2001; 52(5): 501–509. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakahashi K, Nagai T, Kamei H, et al.: Neural circuits containing pallidotegmental GABAergic neurons are involved in the prepulse inhibition of the startle reflex in mice. Biol Psychiatry. 2007; 62(2): 148–157. PubMed Abstract \| Publisher Full Text\n\nTakei H, Song L, Ebihara K, et al.: Histaminergic effects on the frequency of repetitive spike firing in rat insular cortex. Neurosci Lett. 2012; 518(1): 55–59. PubMed Abstract \| Publisher Full Text\n\nTasaka K, Chung YH, Sawada K, et al.: Excitatory effect of histamine on the arousal system and its inhibition by H1 blockers. Brain Res Bull. 1989; 22(2): 271–275. PubMed Abstract \| Publisher Full Text\n\nThomsen MS, Christensen DZ, Hansen HH, et al.: alpha(7) Nicotinic acetylcholine receptor activation prevents behavioral and molecular changes induced by repeated phencyclidine treatment. Neuropharmacology. 2009; 56(6–7): 1001–1009. PubMed Abstract \| Publisher Full Text\n\nTobia MJ, Iacovella V, Hasson U: Multiple sensitivity profiles to diversity and transition structure in non-stationary input. NeuroImage. 2012; 60(2): 991–1005. PubMed Abstract \| Publisher Full Text\n\nvan Wamelen DJ, Shan L, Aziz NA, et al.: Functional increase of brain histaminergic signaling in Huntington’s disease. Brain Pathol. 2011; 21(4): 419–427. PubMed Abstract \| Publisher Full Text\n\nVinogradov S, Solomon S, Ober BA, et al.: Do semantic priming effects correlate with sensory gating in schizophrenia? Biol Psychiatry. 1996; 39(9): 821–824. PubMed Abstract \| Publisher Full Text\n\nWang F, Kalmar JH, He Y, et al.: Functional and structural connectivity between the perigenual anterior cingulate and amygdala in bipolar disorder. Biol Psychiatry. 2009; 66(5): 516–521. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWiesendanger R, Wiesendanger M: The corticopontine system in the rat. I. Mapping of corticopontine neurons. J Comp Neurol. 1982; 208(3): 215–26. PubMed Abstract \| Publisher Full Text\n\nYadon CA, Bugg JM, Kisley MA, et al.: P50 sensory gating is related to performance on select tasks of cognitive inhibition. Cogn Affect Behav Neurosci. 2009; 9(4): 448–458. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYamada S, Harano M, Tanaka M: Dopamine autoreceptors in rat nucleus accumbens modulate prepulse inhibition of acoustic startle. Pharmacol Biochem Behav. 1998; 60(4): 803–808. PubMed Abstract \| Publisher Full Text\n\nYee BK: Cytotoxic lesion of the medial prefrontal cortex abolishes the partial reinforcement extinction effect, attenuates prepulse inhibition of the acoustic startle reflex and induces transient hyperlocomotion, while sparing spontaneous object recognition memory in the rat. Neuroscience. 2000; 95(3): 675–689. PubMed Abstract \| Publisher Full Text\n\nYoung KA, Randall PK, Wilcox RE: Startle and sensorimotor correlates of ventral thalamic dopamine and GABA in rodents. Neuroreport. 1995; 6(18): 2495–2499. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5257", "date": "21 Jul 2014", "name": "Stan Leung", "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\nSkefos et al. reported that administration of H1 receptor antagonist pyrilamine with NMDA receptor antagonist MK-801 (daily for 4 weeks) reversed the impairment of auditory prepulse inhibition (PPI) caused by dizocilpine (MK-801) administration alone. Pyrilamine treatment resulted in an increase in α7 nicotinic receptor binding in the insular cortex, which also correlated with PPI improvement.A change in α7 nicotinic receptor binding in the insular cortex after pyrilamine administration, in correlation with PPI improvement, is a novel result. α7 nicotinic receptor binding in the anterior cingulate cortex also correlated with PPI, but did not change with pyrilamine treatment. The participation of histamine in cholinergically mediated arousal forebrain circuits is known for some time and nicotinic agonism, and possibly an increase in α7 nicotinic receptor binding in the insular cortex, may improve PPI. However, how administration of an H1 antagonist resulted in increased α7 nicotinic receptor binding is not known. The authors' inference of a \"wide distribution of networks capable of influencing PPI of startle\" is likely correct. While a direct participation of insular cortex in PPI is not excluded, correlation of binding with PPI may suggest two separate and parallel events after pyrilamine treatment, without cause and effect relation. The title is appropriate. The abstract reads well, but should state chronic pyrilamine treatment. The paper is generally well written. However, some methods and results, and sample sizes (n) should be presented more clearly or explicitly before indexing.The methods included testing PPI every week, and the result presented in Fig. 2 was apparently for the final PPI test at 4 weeks. Were similar results observed for the PPI tests at earlier times? Did the groups differ in their startle response (without prepulse)? To be clear, the Results section should state “pyrilamine with dizocilpine” administration (not just pyrilamine) “significantly attenuating the dizocilpine-induced PPI impairment.” Did pyrilamine alone in Fig. 2 significantly alter PPI as compared to saline alone? The equation for PPI should have x100% on the right side. The Results and figures/legends did not state the sample sizes, and in some cases, the inclusion criteria were not clear. There were apparently 4 groups as indicated in Fig. 1 (control, saline administered; pyrilamine alone; MK-801 alone; MK-801+ pyrilamine). In the Methods, the number of rats (n=9) was only stated for the last two groups. (The data sets do clarify that the other two groups also had n=9). Do data points of “no pyrilamine” Pyr(-) in Fig. 3 correspond to the MK-801 group without pyrilamine or also include the saline control group? Or does the Pyr(+) group include data with and without MK-801? Did Pyr have the same effect with and without MK-801? Similar questions can be asked for other figures, and the authors should be explicit as to which groups were included. Currently, Fig. 3 appears to have 13-14 points for each of the Pyr (-) and Pyr (+) groups, and ~24 points for the correlation in Fig. 4 [Fig. 5: 15 & 17 points, and Fig. 6, 30 points]. These “n” are unclear to the reader, and should be stated in the figure or figure legend. If the authors had used more than one point for each rat, this has to be explained. Statistics should include degrees of freedom (df) or sample size (n); e.g., 37 ± 19% (give n), and R2 statistic should include df.", "responses": [] }, { "id": "7535", "date": "10 Feb 2015", "name": "David Reser", "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 an interesting attempt to link histaminergic changes in PPI during H1 receptor blockade with pyrilamine in the presence of the NMDA receptor antagonist dizocilpine (MK-801) with changes in H1 binding measured by autoradiography. However, I was left with several questions regarding the study population and presentation of the data:The study was conducted on 36 female rats, with no rationale provided for the exclusion of males. It would be helpful to consider in the discussion how this choice may impact PPI changes in light of steroid receptor interactions with both H1 and a7-nicotinic receptors. The association between decreased H1 binding and increased PPI is presented in the results as \"R2 = -0.15\", which is not possible, so I assume that refers to the Pearson's R value. Thus the correlation was very weak, and although they report a significant result (p=0.05), that is hardly a strong effect. Examination of figure 4 suggests that the population may have divided into responders and non-responders, at least for the PPI measured at 77dB. Approximately 9 of the data points in figure 4 show negligible PPI even at the lowest measured H1 binding, which suggests that there may be other factors affecting this relationship. Cluster analysis of the data might be revealing in this context. No rationale is provided for testing only a single dose of pyrilamine, which is somewhat problematic given the very small difference between pyrilamine (+) and (-) animals in figure 3, which compares H1 binding in the anterior cingulate between treated and untreated animals. The description of this result as a \"nearly significant trend\" (p=0.058) is also unconvincing, especially as there is a possible outlier in the pyr (-) group. A Dixon test or similar approach could be applied to determine if this is a true outlier, in which case there would be virtually no difference between groups. It would be helpful to include in the discussion some consideration of the specificity of pyrilamine for the H1 receptor, and the possible consequences of the dosage applied on other receptor subtypes or classes. A number of recent reports are available which discuss the interactions of pyrilamine with nicotinic receptors in other contexts (e.g. Sadek et al., 2015; Kim et al., 2014), and these should be cited and discussed with respect to the present findings.", "responses": [ { "c_id": "1220", "date": "12 Feb 2015", "name": "David Reser", "role": "Reviewer Response", "response": "The more I look at this, the more uncomfortable I am with the choice of analysis methods applied to the comparison of receptor binding with PPI change. In figures 6 and 7, nearly all of the subjects with non-zero PPI have y-axis values above the line of correlation, due to the absence of PPI in all of the other subjects. In a sense, virtually all of the data points in this correlation analysis are \"residuals\", located well away from what should approximate a first-order line of best fit through the population. The correlation measures are thus biased to the point of uselessness, since they really reflect the middle values between populations of apparent responders and non-responders. I strongly recommend that the data in these figures, along with the data from fig. 4 (as called out in the review above) be re-analyzed, possibly with the assistance of a statistician, as these populations are quite likely clustered, and therefore not suitable for a simple correlation analysis." } ] } ]	1	https://f1000research.com/articles/3-136
https://f1000research.com/articles/3-100/v1	02 May 14	{ "type": "Short Research Article", "title": "A promising approach to scale up health care improvements in low-and middle-income countries: the Wave-Sequence Spread Approach and the concept of the Slice of a System", "authors": [ "M. Rashad Massoud", "Nana Mensah-Abrampah", "M. Rashad Massoud" ], "abstract": "There are several examples of successes in improving health care. However, many of these tend to remain limited to the sites they were originally developed in. There are fewer examples of successes that have spread to the remainder of the systems they were developed in. This article discusses the wave-sequence approach, which is a type of spread focusing on spreading in a systematic, sequential manner, using spread agents—people who participated in the original demonstration sites—to spread to other parts of the systems. The paper also discusses the associated concept of the “slice” of a system. In addition, we describe a phenomenon related to the rate of adoption that we are seeing in the wave-sequence approach.", "keywords": [ "There are several examples of successes in improving health care. However", "many of these successes are limited to the sites where they were originally developed in", "a phenomenon referred to as “islands of excellence”. There are fewer examples of success that spread to the remainder of the system where the improvement was originally developed. “Spread” (or “scale-up”) is the science of taking a local improvement (e.g. the implementation of an intervention", "the redesign of a process or system) that has produced a better result than the previous method", "and actively disseminating it across a system1. There are many ways to “spread”", "including", "but not limited to", "natural diffusion", "extension agents", "emergency mobilization", "collaborative improvement", "virtual collaborative methods", "campaign spread", "the wave-sequence approach", "and hybrid models1–3. Notable examples of spread include the 100", "000 Lives Campaign of the Institute of Health Care Improvement", "the Quality Assurance Project (QAP) in Russia", "which was funded by the United States (US) Agency for International Development (USAID)", "the USAID Health Care Improvement Project (HCI) in Afghanistan4", "and the Project Fives Alive! in Ghana5", "which was funded by the Bill and Melinda Gates Foundation. This article focuses on only one of these approaches: the wave-sequence spread approach. It also discusses the associated concept of the “slice” of a system and describes a phenomenon related to the rate of adoption that we are seeing in the wave-sequence approach." ], "content": "Introduction\n\nThere are several examples of successes in improving health care. However, many of these successes are limited to the sites where they were originally developed in, a phenomenon referred to as “islands of excellence”. There are fewer examples of success that spread to the remainder of the system where the improvement was originally developed. “Spread” (or “scale-up”) is the science of taking a local improvement (e.g. the implementation of an intervention, the redesign of a process or system) that has produced a better result than the previous method, and actively disseminating it across a system1. There are many ways to “spread”, including, but not limited to, natural diffusion, extension agents, emergency mobilization, collaborative improvement, virtual collaborative methods, campaign spread, the wave-sequence approach, and hybrid models1–3. Notable examples of spread include the 100,000 Lives Campaign of the Institute of Health Care Improvement; the Quality Assurance Project (QAP) in Russia, which was funded by the United States (US) Agency for International Development (USAID); the USAID Health Care Improvement Project (HCI) in Afghanistan4, and the Project Fives Alive! in Ghana5, which was funded by the Bill and Melinda Gates Foundation. This article focuses on only one of these approaches: the wave-sequence spread approach. It also discusses the associated concept of the “slice” of a system and describes a phenomenon related to the rate of adoption that we are seeing in the wave-sequence approach.\n\n\nThe wave-sequence approach\n\nThe wave-sequence approach is a type of spread that focuses on spreading improved care delivery to other parts of the system. An example of health care delivery system which has yielded better results in a systematic, sequential manner, using spread agents—people who participated in the original demonstration sites—is the implementation of the Active Management of the Third Stage of Labor (AMTSL). Demonstration sites might be hospitals and several clinics, and people from those facilities may assist in spreading the intervention to the remaining facilities in a district, (administrative) department, or city. The term “wave” reflects the fact that this method of spread occurs both sequentially and in an increasingly larger section of the same health care system (perhaps the whole system). Wave-sequence spread is used when it is not possible to cover the whole system all at once. For simplicity, this article uses “wave 1” to refer to the demonstration phase and “wave 2” onwards to refer to the subsequent spread phase; however, some experts use “demonstration” for the first phase and follow with wave 1, etc. Also for simplicity the geographic and administrative unit mentioned in this article refers to the aim of achieving coverage at a whole country level, but the wave-sequence approach can be used at any level: facility, community, district, and on up. It has also been used for spread from an initial ward in large hospitals to the remainder of the wards in those hospitals.\n\nThe yellow-colored area in Figure 1A represents the full geographic scale we want to spread to, covering the entire country. This would have administrative sub-divisions, such as districts, provinces, or states (this article uses “districts”). Figure 1B presents each of those districts as one of the “petals” of the “flower”. Unlike a real flower, the center of the flower is also referred to as a petal. Each district would have a central city, represented by a red dot.\n\n(A) Full geographic scale that we want to spread to, covering the entire country. (B) Administrative sub-divisions such as districts, provinces, or states represented as petals in a country. (C) Blue dots representing health facilities and community structures that provide service(s) in a country. (D) Movement of spread agents from their original facilities to facilities in other slices.\n\nThroughout the country, health facilities and community structures provide the service(s) for the outcomes we want to improve. These structures are represented by the blue dots in Figure 1C. For the most part in global health, we are dealing with nested systems. These have common administrative structures (usually a Ministry of Health, referral systems, etc.), which are represented by the lines between the blue dots. The green triangles in this figure are described below as “slices” of the system.\n\nThe wave-sequence approach is used when we cannot reach the full scale all at once. Therefore, we start with a selection of facilities—those in the wave 1 slices—and later, the staff of the facilities in wave 1 slices will spread the intervention to the remainder of the districts in the system.\n\n\nThe “Slice of a System” concept\n\nIt is best to select a slice of the system in each of the system’s geographic or administrative sub-divisions. A slice (green triangles in Figure 1C) is the set of facilities and community structures (e.g., a community health worker who may work from home) that provide the service that we want to improve and cover a subset of population in each sub-division. The slice has a number of elements ranging from community-level structures to primary care facilities and hospitals including referral facilities. All the structures in a slice would include all the services affecting the outcome of interest (e.g., reducing maternal mortality). This is because patients receive care along the whole continuum of delivery. We would include all those facilities and community structures in the slice to start with. If this were not possible, we would take enough facilities to represent all the different types of facilities in the slice. We work on the initial improvement in these slices, engaging representatives of care providers and their leaders in developing the improvements we intend to make. If for any reason (financial, political, geographic), we cannot take a slice of the system in each sub-division; we take slices in some of the regions but make the selection to the extent possible such that they represent the different settings in the whole target area.\n\nAs these improvements are achieved during wave 1, we watch carefully for the providers who are most engaged in the work and produce the best results in the initial set of slices. Towards the end of wave 1, we ask each such provider whether he/she would like to participate in the spread wave(s). For those who accept, we work with them and equip them to become “spread agents” for the subsequent waves that will address the remainder of the system (their spread from their original facilities to facilities in other slices are represented by black lines in Figure 1D). What we look for in the providers is not just the technical skills and skills in quality improvement, but also the ability to teach and coach others in this work. From the beginning of the project, we also make arrangements with the health authorities so that some of their staff will be permitted to play the role of providers or spread agents in the subsequent waves. Arrangements are also made with the health authorities so that they visibly lead the spread in the subsequent waves. Other issues that need to be addressed in the subsequent waves include integrating the data management, communications, meetings, and events. Managing these logistical issues will foster the spread of the intervention into the regular management of the health system.\n\n\nExamples of the wave-sequence approach: Russia and Afghanistan\n\nStarting in 1998, the USAID QAP worked with their counterparts at the Institute of Public Health in Moscow and the Ministry of Health in Russia on improving care in selected demonstration sites in two of Russia’s 89 oblasts and territories: Tula and Tver Oblasts. The technical areas covered were hypertension (HT), pregnancy-induced hypertension (PIH) and neonatal respiratory distress syndrome (NRDS)—within a couple of years, significant improvements in the quality of care were achieved in all three technical areas. For example, case fatality from NRDS fell by 64% (Figure 2 and Figure 3). The key changes made to the NRDS care delivery system were to re-organize it into four components:\n\n1. Improve competencies in neonatal resuscitation for pediatricians, obstetricians, midwives and nurses at the point of delivery of newborns;\n\n2. Provide a neonatal transport system consisting of four equipped vehicles;\n\n3. Strengthening the neonatal intensive care units (NICUs) in Tver City, which already existed and were equipped with neonatal ventilation capability. These units received neonates suffering from respiratory distress (a condition requiring referral from the other sites we were working in) from all over Tver Oblast;\n\n4. Implementing policy level changes that facilitated the referral of neonates to higher-level facilities.\n\nAs a result of the neonatal transport system, the percentage of neonates referred to the NICU with abnormally low body temperature was significantly reduced. This improves the survival of newborns with respiratory distress syndrome.\n\nAs a result of the improvement, fewer neonates died in the first week of life.\n\nThe work in Russia used the collaborative improvement method as part of the wave-sequence approach. Collaborative improvement brings together different teams working on a common aim and associated indicators to change and improve processes of care delivery3. Each team collects data on a common set of core indicators to measure whether changes introduced to the system are resulting in improvement. Those changes resulting in improved outcomes are then spread to all the sites involved in the collaborative approach. In itself, collaborative improvement is regarded as a spread approach because it involves a large number of sites working together to improve care. When providers who excelled from the original sites using collaborative improvement become the spread agents to the remainder of the system, this is the wave-sequence spread approach.\n\nAfter witnessing significant improvements in a relatively short time and the motivating effect that applying improvement methods had on the health providers, the participants in the wave 1 sites turned their attention to how these successes and the use of the improvement methods could be spread further. The goal (set jointly by the USAID QAP, Institute of Public Health in Moscow, Ministry of Health in Russia, and Tula and Tver Oblasts Health Department) for scale up was set from five to 43 hospitals for NRDS, from three to 40 hospitals for PIH (all Tver Oblast), and from five to 442 polyclinics for HT in Tula Oblast6.\n\nThe project ultimately had two waves in Tver Oblast and several overlapping waves in Tula, which involved a larger number of facilities. Additional financial resources to support the spread to all of Tula and Tver Oblasts did not come through to QAP despite all good intentions. However, the leadership of both oblasts decided to pursue spread by adding their own resources to QAP’s in order to support this sizeable effort. This culminated in a re-planning of the spread effort with a new design element: The providers who excelled in wave 1 would serve as spread agents to the reminder of their oblast. To maximize their likelihood of success, QAP and the leadership of Tula and Tver Oblasts built their capacity as spread agents. The spread strategy was designed around six concepts7:\n\n1. The spread agents would come from within the system, not from outside. Specifically, members of the quality improvement teams from wave 1 would act as spread agents in the spread waves.\n\n2. Roles and responsibilities were clearly defined. Each spread team had at least one person with strong skills in improvement and another with strong skills in the technical content (NRDS, PIH, or HT).\n\n3. Spread teams were equipped with the competencies needed to effectively perform their duties as spread agents. They were trained in technical content, improvement methods and coaching techniques, so they had not only the know-how, but also could teach and coach others in it.\n\n4. In order to have a mechanism where different spread teams could share experiences and develop hypotheses, spread agents and their teams would join the regular learning sessions that were part of the collaborative improvement method. These sessions are held every six to eight weeks during a collaborative improvement. Representatives of all the facilities participating in that improvement effort attend these sessions3; they offer participants opportunities to discuss experiences, air problems and harvest solutions, and make plans for further improvement efforts when they return to their facilities7.\n\n5. The Oblast Health Authority visibly led the spread effort. Its monthly meetings (“Kollegia” Meetings) became the channel for managing, reporting, communication and problem solving for the spread effort.\n\n6. External technical assistance was provided primarily to the health authorities and the spread agents. The QAP technical assistance team met regularly with the Oblast Health Authority to review the status of the program and to problem solve. A special series of meetings, the “Masters Seminar,” was held in order to enhance the capacity of the spread agents and share learning between them on an on-going basis.\n\nFigure 4 shows the reduction in early neonatal mortality, neonatal mortality and infant mortality during the demonstration phase, spread phase and for six years after the end of USAID QAP’s technical assistance to Tver Oblast. Significant improvements in mortality were achieved and sustained beyond the life of the project. This data comes from Tver Oblast Health Authority surveillance system. The improvement and scale up method used targeted enhancing the capabilities of the staff in Tver Oblast in continually improving care.\n\nThe reduction in early neonatal mortality, the neonatal mortality and the infant mortality during the demonstration phase, spread phase and for six years after the end of USAID QAP’s technical assistance to Tver Oblast are shown. Significant improvements in mortality were achieved and sustained beyond the life of the project. This data comes from Tver Oblast Health Authority surveillance system.\n\nIn 2009, the Afghanistan Ministry of Public Health (MoPH) asked HCI (the follow-on to QAP) to institutionalize improvement methods in Afghanistan. Reducing maternal and newborn mortality and morbidity was chosen as the improvement priority. The improvement initiative initially took place in two provinces, Balkh and Kunduz, together with a referral hospital in Kabul city. The decision to scale up from two provinces and one hospital in wave 1 to four provinces in wave 2 was part of the initial project design. Slices of the system were selected in each of the three sub-divisions and consisted of 26 facilities: 10 facilities in Balkh, 15 in Kunduz, and the referral hospital in Kabul. These 26 facilities constituted wave 1.\n\nThroughout this wave, the collaborative improvement approach was used. Three types of collaborative improvement team were formed:\n\n1. Provincial facility-based collaborative teams,\n\n2. Provincial community collaborative teams, and\n\n3. Kabul Maternity Hospital collaborative teams.\n\nWith the help of provincial facility-based and community collaborative teams, HCI prioritized a package of high-impact interventions that was introduced to the selected health facilities. The first phase package focused only on antenatal care, the active management of the third stage of labor, essential newborn care, and immediate post-partum care for the mother. The second phase package addressed antenatal screening for complications, management of obstetric and newborn complications (including eclampsia, sepsis, and newborn asphyxia), and post-partum family planning9. These high-impact interventions were known as the change package.\n\nDuring wave 2, facility-based quality improvement teams were established at each participating health facility and trained in quality improvement methodology and how sites achieved improvements during wave 1. The providers who excelled in first wave did this. The second wave followed the same pattern, drawing on the change package of evidence-based practices that had been developed from the health facilities.\n\nIn Afghanistan, three waves were conducted. Wave 2 focused on spread within the provinces. Once the change package from wave 1 was finalized, the aim was to engage three additional health facilities in each learning session. However, demand for rapid spread was intense yet funding was limited. Consequently, six facilities were added in each wave. Wave 3 focused on spread to other provinces. Once wave 1 was seen to attain a remarkable level of results, spreading the program to other provinces became a priority. In this case, HCI staff served as change agents using the title “provincial coordinator” (PC). PCs were trained and deployed to the original demonstration sites and then became the spread agents for the new provinces. The change package from wave 1 was adapted to fit the needs of the new facilities in wave 2. Spread between major hospitals was part of wave 3. Initially, the PCs started with two major hospitals. After a year, three major provincial hospitals were added. In the second year, five provincial hospitals were added. In total, 11 major hospitals participated in the hospital collaborative improvement.\n\nScale-up was designed around using local resources and capacity in Afghanistan. Leadership came from within the MoPH of Afghanistan. HCI used local capacity to develop provincial quality improvement teams who developed a context-specific change package, led the improvement, and introduced the successes from wave 1 to the second and third waves. Collaborative improvement approaches and learning prompted the rapid dissemination of successful practices to achieve results.\n\nBy the end of 2012, the maternal and newborn health facility collaborative improvement interventions had reached 85 health facilities in the nine HCI-targeted provinces, achieving measurable gains in the quality of maternal and newborn care for an estimated total catchment population of over 1.5 million, about 24% of the population of these provinces.\n\nAn important difference between the Russian and Afghan examples is that in Russia, the providers who excelled from the wave 1 site were health providers who led the spread to other sites in the subsequent waves. In Afghanistan, due to local circumstances, it was not possible to use the providers who excelled from wave 1 as spread agents. Instead, HCI staff who were host-country nationals and seen as peers by health staff, worked with the wave 1 sites and undertook the role of spreading the wave 1 experience to the spread sites. In both cases, the common factors were intricate knowledge of the work in wave 1 and being a host-country national with knowledge of the local context.\n\nFigure 5 is a time-series chart from Afghanistan that presents the results of the collaborative improvement that sought to reduce maternal deaths by introducing active management of the third stage of labor. This practice has three services, all delivered immediately after birth to prevent hemorrhage: provision of oxytocin, cord traction, and uterine massage. The graph covers the period June 2009–December 2012, with the results of wave 1 shown in green, wave 2 in blue, and wave 3 in red. Note that the lines for waves 2 and 3 are much steeper than that for wave 1, indicating that the rate of provider adoption of this intervention was much faster than wave 1, a phenomenon that is discussed next.\n\nThe rate of adoption of AMTSL in waves 2 and 3 was significantly faster than in the initial wave.\n\n\nRate of adoption in subsequent waves\n\nIt is commonly believed that it is difficult to replicate the results of a well-conducted demonstration or pilot project at a much larger scale and still achieve similarly good results. The experience of many global health care projects has been that as they are scaled up, fidelity to the intervention and its results were not good as in the initial demonstration. This conversation emerged at the end of the scale-up planning session in Russia. Stakeholders agreed “if we can get half as good results in the scale-up as we did in the demonstration, we would be happy”.\n\nSurprisingly though, provider adoption happened faster and more easily and achieved even better results. Implementation of the second and subsequent waves also had fewer problems: spot checks confirmed that the adoption was happening with fewer problems. The ease with which the changes could be transferred and applied by new facilities during subsequent waves can possibly be attributed to the “homophily” factor: the degree to which pairs of individuals who interact are similar in certain attributes, such as beliefs, education, social status and the like2. The spread was conducted by peer providers who had implemented the changes and obtained the improved results in similar facilities in the same context.\n\nWe have since seen this faster/easier/better phenomenon in other places and other interventions including AMTSL in Ecuador10, tuberculosis (TB) and HIV in Uganda, Prenatal Care in Guatemala and Integrated Management of Childhood Illness in Niger (Figure 6–Figure 9). Due to the similarity in context from wave 1 to subsequent waves in these countries, relatively few adaptations were made to the change package developed in wave 1. This rate of adoption phenomenon warrants further research11. Please note that these graphs were chosen to reflect the rate of spread/acceptance/adoption we are seeing with the wave sequence spread approach.\n\nThe rate of adoption of screening for TB in HIV patients was much faster in wave 2 than wave 1.\n\nThe rate of adoption of triage as part of IMCI was much higher in wave 2 than wave 1.\n\nThe rate of adoption of compliance with AMTSL was much faster in wave 2 and 3 than in wave 1.\n\nThe rate of adoption of compliance with Prenatal Care Criteria in wave 2 was much faster than in wave 1.\n\nThe first graph in Figure 6 shows HCI experience in Uganda with regard to an improvement effort to screen people who were HIV-positive for TB. While both waves reached an 80% adoption [spread/acceptance] rate, wave 2 started at a far lower percentage (albeit with fewer clients) and climbed very steeply compared to wave 1. The second graph shows HCI experience in Niger relative to the adoption of the integrated management of childhood illness. While the wave 1 line seems to stagger up the chart, that for wave 2 rises expeditiously. Again, in Ecuador, the graph shows that waves 2 and 3 moved much more quickly than wave 1, and wave 3 was faster that wave 2. Finally, in Guatemala’s efforts to improve pre-natal care, the wave 1 sites fell from a starting point of almost 80% to a low of 63% before rising to just above 90%. Wave 2, however, made a more rapid rise from 18% to 84%.\n\n\nConclusions\n\nIn situations where full-scale adoption of an innovation cannot be reached all at once, the wave-sequence approach has proven to be useful. We believe the reason why an innovation can spread from one set of sites to others is due to the fact that once something has proved to be successful in one set of sites, providers in the following set of sites are more likely to believe it will succeed in theirs. While it has been common for pilot projects to fizzle out when spread to other sites, we believe the successes of the wave sequence approach are based on engaging the host country national staff who have experienced the demonstration wave to champion the spread of implementing the specific changes to others. Testing on a small scale, in the first wave also makes it easier to overcome any managerial, policy-level or other challenges.\n\nIt is critical to get results in wave 1 before initiating new waves. Such results give credibility to the proposed changes, giving implementers confidence in the change package. Another important aspect is that though wave 1 can rely on an external development agency, subsequent waves should be much more integrated into the existing national structure of the host-country.", "appendix": "Author contributions\n\n\n\nM. Rashad Massoud participated in the design and implementation of the work in Russia. He also contributed to the design of the work in Afghanistan. M. Rashad Massoud and Nana Mensah-Abrampah contributed equally to the development and writing of this manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe Quality Assurance Project (QAP) was funded by the U.S. Agency for International Development (USAID), under Contract Number HRN-C-00-96-90013. The project is managed by University Research Co., LLC (URC). The USAID Health Care Improvement Project is funded by the American people through the United States Agency for International Development and its Bureau for Global Health. The project is managed by University Research Co., LLC (URC) under the terms of Contract Number GHN-I-03-07-00003-00.\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\nRussian Federation Ministry of Health: VI Starodubov, RR Khalfin, and II Rumina. Central Public Health Research Institute (CPHRI) in Moscow: YM Komarov, AV Korotkova, and NV Surova. Tver Oblast Department of Health: ANZlobin, OV Chernobrovkina, LK Samoshkina, TP Dmitrievna, and TP Gviniashvilli. Tula Oblast Department of Health: EI Chernienko and VA Melnikov.\n\nAfghanistan: M Rahimzai, E Yaqubi, A Amiri, KM Amiri, AK Hekmati, RH Salem, F Hakimi, M Ahmadzai, and I Sahak.\n\n\nReferences\n\nMassoud MR, Donohue KL, McCannon CJ: Options for Large-scale Spread of Simple, High-impact Interventions. Technical Report. Bethesda, MD: USAID Health Care Improvement Project, University Research Co., LLC (URC). 2010. Reference Source\n\nRogers EM: Diffusion of Innovations. New York: Free Press. 2003. Reference Source\n\nFranco LM, Marquez L: Effectiveness of collaborative improvement: evidence from 27 applications in 12 less-developed and middle-income countries. BMJ Qual Saf. 2011; 20(8): 658–65. PubMed Abstract \| Publisher Full Text\n\nMassoud MR, Askov K, Reinke J, et al.: A Modern Paradigm for Improving Healthcare Quality. QA Monograph Series 1(1). Bethesda, MD: USAID Quality Assurance Project, URC. 2001. Reference Source\n\nTwum-Danso NA, Akanlu GB, Osafo E, et al.: A nationwide quality improvement project to accelerate Ghana’s progress toward Millennium Development Goal Four: design and implementation progress. Int J Qual Health Care. 2012; 24(6): 601–11. PubMed Abstract \| Publisher Full Text\n\nMassoud MR, Nielsen GA, Nolan K, et al.: A Framework for Spread: From Local Improvements to System-Wide Change. IHI Innovation Series white paper. Cambridge, Massachusetts: Institute for Healthcare Improvement. (Available at www.IHI.org). 2006. Reference Source\n\nBerwick D: Lessons from developing nations on improving health care. BMJ. 2004; 328(7448): 1124–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCannon CJ, Berwick DM, Massoud MR: The science of large-scale change in global health. JAMA. 2007; 298(16): 1937–9. PubMed Abstract \| Publisher Full Text\n\nRahimzai M, Amiri M, Burhani NH, et al.: Afghanistan’s national strategy for improving quality in health care. Int J Qual Health Care. 2013; 25(3): 270–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHermida J, Salas B, Sloan NL: Sustainable scale-up of active management of the third stage of labor for prevention of postpartum hemorrhage in Ecuador. Int J Gynaecol Obstet. 2012; 117(3): 278–82. PubMed Abstract \| Publisher Full Text\n\nHeiby J: The use of modern quality improvement approaches to strengthen African health systems: a 5-year agenda. Int J Qual Health Care. 2014; 26(2): 117–23. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5011", "date": "05 Jun 2014", "name": "John Ovretveit", "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\nEnglish in abstract – I made slight changes - below is the suggested replacement:“There are several examples of successes in improving health care. However, many of remain limited to the sites at which they were originally developed. There are fewer examples of successful spread of the improvement more widely in side or outside the health systems within which they were developed. This article discusses a wave-sequence approach to spread or scale up, which enables take up of the improvement in a systematic and sequential way, using “spread agents” —people who participated in the original demonstration sites. The paper also discusses the concept of the “slice” of a system which is useful for thinking about spread and considers a phenomenon related to the rate of adoption which we have observed in this wave-sequence approach.” The title is appropriate for the article, which is well constructed and clear. The methods used as accepted methods for this type of study, and the analyses adequate. The conclusions appear sensible and balanced.", "responses": [ { "c_id": "865", "date": "17 Jun 2014", "name": "Nana Mensah-Abrampah", "role": "Author Response", "response": "Dear John,Thank you for your review. We very much appreciate your positive comments. We also appreciate your revisions to the abstract. We agree and have modified the abstract accordingly.Kind regards,Nana and Rashad" } ] }, { "id": "5122", "date": "16 Jun 2014", "name": "Lisa M. Schilling", "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 appropriately addresses both the social system needs for spread using \"spread agents\" and a core technical need around identifying the common unit or slice that is involved in adoption and spread. This is important information for healthcare professionals to understand. It also explains why a \"wave\" approach to spread could enhance adoption over the pure diffusion method. The article does not address the \"characteristics\" of the intervention itself, complexity for example, nor how adaptation to sites that are somewhat different affects the success of spread. I look forward to subsequent publications that may address such topics.", "responses": [ { "c_id": "864", "date": "17 Jun 2014", "name": "Nana Mensah-Abrampah", "role": "Author Response", "response": "Dear Lisa,Thank you for your thoughtful review. We very much appreciate your positive comments. We agree that the article does not address the characteristics of the intervention itself. Given the allowable size of the article, we were not able to discuss several other issues that we would have liked to add. Your idea about subsequent publications that may address this topic is a great one. We look forward to future publications that deal with this.Thank you,Nana and Rashad" } ] } ]	1	https://f1000research.com/articles/3-100
https://f1000research.com/articles/3-134/v1	24 Jun 14	{ "type": "Research Note", "title": "Additive effects of β-adrenergic and cytokine signaling on lipolytic activation", "authors": [ "Nicholas Card", "William S. Garver", "Robert A. Orlando", "Nicholas Card", "William S. Garver" ], "abstract": "Obesity often leads to increased systemic inflammation which is now thought to play a causative role in the development of atherosclerotic disease and insulin resistance. This inflammatory response originates within large adipose tissue depots and is initiated by classically activated macrophages that infiltrate the tissue from the circulation. The large number of macrophages residing in obese adipose tissue leads to significant increases in interleukin-6 (IL-6) and tumor necrosis factor-α (TNFα) secretion; achieving levels sufficient to elevate circulating plasma concentrations. These cytokines activate potent signals to initiate lipolysis, to release free fatty acids from triacylglycerol stores and contribute to hyperlipidemia in obese individuals. Obese adipose tissue responds to normal β-adrenergic and glucagon stimuli to recover from negative energy balance by inducing lipolysis. However, it is not clear what quantitative influence additional lipolytic stimulation by IL-6 and TNFα has on normal β-adrenergic activity. Although, β-adrenergic and cytokine signaling activate separate pathways for lipolytic activation, it is undefined whether the effects of multiple signaling events on lipolysis are additive or coincident. To clarify this issue, we measured lipolytic activity in 3T3-L1-derived adipocytes stimulated by a β-adrenergic agonist (isoproterenol), IL-6 or TNFα individually and in combinations as co- and tri-stimulation. Treatment of adipocytes with isoproterenol and either IL-6 or TNFα as co-stimulants increased lipolytic activation by approximately the sum of the individual ligands, suggesting contributions from two independent pathways. Co-stimulation with IL-6 and TNFα provided slightly more than an additive response indicating signaling contributions from independent and common pathways. Tri-stimulation resulted in the largest level of lipolytic activation with a value approximate to adding isoproterenol stimulation to a combined treatment of IL-6 and TNFα. The additive nature of cytokine signaling to β-adrenergic activity suggests its therapeutic inhibition will prevent excessive lipolysis, yet minimally interfere with maintaining normal responses to varying energy demands.", "keywords": [ "adipocyte", "obesity", "cytokine", "β-adrenergic", "TNFα", "IL-6", "lipolysis" ], "content": "Introduction\n\nObesity often leads to increased systemic inflammation which is now thought to play a causative role in the development of atherosclerotic disease and insulin resistance1–4. Increasing adiposity due to excessive weight gain sets up a chronic inflammatory response within adipose tissue which is promoted by recruitment and infiltration of classically activated, type-1 macrophages from the circulation5. Although the reasons for sustained adipose inflammation remain unclear, the large number of macrophages residing in obese adipose tissue leads to significant increases in secretion of interleukin-6 (IL-6) and tumor necrosis factor-α (TNFα); achieving levels sufficient to elevate circulating plasma concentrations6. IL-67 and TNFα8,9 are multifunctional cytokines that have well-established roles in inflammatory responses. In adipose tissue, these cytokines share a common activity in delivering potent signals for stimulation of lipolysis10–13. Increased lipolytic activity in obese adipose tissue increases free fatty acid (FFA) flux into the circulation. From a lipocentric view, elevated plasma levels of non-esterified fatty acids (NEFA) leads to increased amounts of atherogenic lipoproteins in the circulation resulting in unmanaged hyperlipidemia14, often accompanied by reduced systemic insulin responsiveness15,16.\n\nThe metabolic actions of IL-6 are diverse. Determination of which actions take precedence is largely dictated by tissue and metabolic context17. For example, IL-6 is an important mediator of the acute phase response which includes hepatic effects to increase glucose output and elevate CRP levels17,18. IL-6 secretion is also increased as a result of exercise19,20, which in turn increases glucose oxidation21 and insulin sensitivity22 in skeletal muscle. IL-6 infusion in humans increased circulating FFA levels11,23, and when adipose tissue or isolated adipocytes were treated with IL-6, lipolytic activity was increased17. In adipocytes, IL-6 binds to a cell surface heterodimer composed of the IL-6 receptor and gp13024,25, and activates two intracellular signaling pathways; the Janus kinase/Signal Transducer and Activator of Transcription (JAK/STAT) pathway, and the p44/42 Mitogen-activated protein kinase (MAPK) pathway17,26.\n\nTNFα is a potent metabolic effector that, in adipocytes, signals primarily through TNFα receptor-127. Intracellular signaling in adipocytes is mediated by p44/42 MAPK and Jun N-terminal kinase (JNK)12,28. Once activated, these pathways induce phosphorylation of perilipin to recruit hormone sensitive lipase for triacylglycerol hydrolysis and the release of FFA, and to downregulate perilipin expression29,30. Perilipin is a phosphoprotein that coats intracellular lipid droplets in adipocytes to maintain minimal lipolytic activity. Phosphorylation of perilipin serves a dual purpose: to release bound CGI-58 to activate adipose triglyceride lipase and to relocate perilipin away from the lipid droplet permitting catalytic access for activated lipases31. However, signaling the sequence of events leading to lipolytic activation through JAK/STAT, p44/42 MAPK and JNK pathways is not the normal physiologic response to increased energy demands requiring the release of fatty acid fuel stores from adipocytes. Normal physiologic activation of lipolysis by β-adrenergic and glucagon signaling during periods of increased systemic energy demands is mediated through heterotrimeric G-protein activation followed by increased intracellular cAMP and protein kinase-A (PKA) activation. Obese adipose tissue is subject to normal β-adrenergic and glucagon stimuli to regulate energy balance; however, this tissue is also subject to additional lipolytic stimuli by IL-6 and TNFα. With the presence of multiple pathways for lipolytic activation, i.e. the normal endocrine/neural pathway and cytokine-mediated pathways, it is unclear whether the effects of multiple signaling events on lipolysis are additive or coincident; that is, do IL-6 and TNFα stimulate lipolytic activities that are in excess of that provided by maximal β-adrenergic activation, or do the pathways activated by these cytokines merge into the downstream β-adrenergic pathway and are they unable to add significantly beyond maximal β-adrenergic activation. To address this question, we have measured lipolytic activity in 3T3-L1-derived adipocytes activated by a β-adrenergic agonist, IL-6 or TNFα, both individually and in combination as co- and tri-stimulation experiments.\n\n\nMaterials and methods\n\nMouse 3T3-L1 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA), and grown and differentiated in complete Dulbecco’s Modified Eagle Medium (DMEM; Irvine Scientific, Santa Ana, CA) with high-glucose and supplemented with 10% fetal bovine serum (FBS; Irvine Scientific, Santa Ana, CA) and incubated at 37°C in a 5% CO2 environment. Twelve-well culture plates (Costar Tissue-culture Treated; Corning, Tewksbury MA) were coated with 1% gelatin (Sigma-Aldrich, St. Louis, MO) prior to cell seeding. After reaching confluency, cells were differentiated into mature adipocytes by incubating with 450 µM 3-isobutyl-1-methylxanthine, 250 nM dexamethasone and 167 nM insulin (Sigma-Aldrich, St. Louis, MO) diluted into DMEM for 3 days, followed by 4 days supplementation with 167 nM insulin alone. Differentiation was confirmed by visual examination (MicrosOpics, IV900 Series inverted microscope) of cells to assess for lipid droplet formation and morphological changes. Prior to each experiment, cells were incubated with complete DMEM without insulin supplementation for a period of 24 hours. Cells were then treated either with isoproterenol (25 nM to 2.5 µM) (Sigma-Aldrich, St. Louis, MO), TNFα (0.1 nM to 25 nM) (Cell Signaling, Danvers, MA) or IL-6 (0.5 nM to 8 nM) (eBioscience, San Diego, CA) for 24 hours to identify an optimal concentration of effector necessary for induction of fatty acid hydrolysis and glycerol release. Once this concentration was determined, subsequent experiments were performed by incubating cells with this identified concentration for 24 hours.\n\nLipolytic determinations were made by quantifying glycerol released into the culture medium following treatments. Culture media were collected and processed using Free Glycerol Reagent according to the manufacturer’s instructions (Sigma-Aldrich, St. Louis, MO). Following the prescribed incubation, the resulting quinoneimine dye in samples was measured by spectrophotometry at 540 nm (Molecular Devices, Spectramax 384 Plus, Sunnyvale, CA). Glycerol amounts in the culture media were determined through comparison with standard curves that were generated by parallel quantification assays using known concentrations of glycerol.\n\nMean values obtained for each experimental treatment group were compared by a one-way analysis of variance (ANOVA) using SigmaPlot 2001 software. Statistical significance is reported when P-values < 0.05.\n\n\nResults and discussion\n\nOur primary objective for this study was to determine if cytokine stimulation (IL-6 or TNFα) heightens lipolytic activity (as measured by glycerol release) over and above what is achieved by normal β-adrenergic signaling. To obtain this quantitative evaluation, we first incubated mature 3T3-L1-derived adipocytes with varying concentrations of either isoproterenol (a β-adrenergic agonist), IL-6 or TNFα individually in order to determine the concentration of ligand that provided maximal lipolytic stimulation. This will ensure that individual ligands will be used at concentrations that maximally activate their respective signaling pathways in co- and tri-stimulation experiments. From our titration study, we have determined that maximal lipolytic stimulation for each ligand is achieved with the following concentrations: 1 µM for isoproterenol, 2 nM for IL-6 and 0.58 nM for TNFα (data not shown).\n\nCo- and tri-stimulation experiments were performed using the ligand concentrations determined above. When adipocytes were incubated with the individual ligands, different levels of lipolytic activation were noted with IL-6 < isoproterenol < TNFα (Figure 1. compare bar 2 < bar 1 < bar 5). These differences are likely due to activation of different signaling pathways which have varying quantitative effects on lipolytic activation. Co-stimulation of adipocytes with isoproterenol and IL-6 resulted in lipolytic activity that was greater than that achieved by stimulation with the individual ligands (Figure 1. compare bar 3 with bars 1 and 2), suggesting parallel activation of different signaling pathways that merge into downstream lipolytic activation. The level of increased lipolytic activation from isoproterenol and IL-6 co-stimulation is approximately the sum of the individual ligands, suggesting an additive effect of contributions from two independent pathways, likely cAMP/PKA and p44/42-JAK/STAT, respectively. When isoproterenol and TNFα were combined, again an additive effect on lipolytic activation was observed (Figure 1. compare bar 6 with bars 4 and 5), similarly suggesting a summing of the effects caused by the activation of two separate signaling pathways, cAMP/PKA and p44/42-JNK, respectively.\n\n3T3-L1 cells were differentiated into mature adipocytes and incubated for 24 hours in the absence of insulin to diminish insulin-dependent anabolic signals. Cells were then incubated for an additional 24 hours in the presence (+) or absence (-) of the indicated concentrations of β-adrenergic agonist (isoproterenol) or cytokine (IL-6 or TNFα) for individual, co- and tri-stimulations. Medium was removed from cells and glycerol was quantified as described in the Materials and methods section. Experimental points were measured in triplicate to determine mean values (shown within bars) and standard deviations (error bars). Data shown are mean values determined using replicate experiments (n = 3). Mean values obtained for each experimental treatment group (1, 2 or 3) were compared by a one-way analysis of variance (ANOVA). Statistical significance is reported when P-values were < 0.05.\n\nIn vivo, obese adipose tissues express both IL-6 and TNFα when they are inflamed and susceptible to concurrent stimulation by both cytokines. To determine the effects of dual cytokine stimulation on lipolysis, adipocytes were incubated with both IL-6 and TNFα. In this case, lipolytic activation was somewhat more than an additive response (Figure 1. compare bar 7 with bars 2 and 5), suggesting that independent, as well as overlapping, signaling pathways were activated. The independent pathways include JAK/STAT for IL-6 and JNK for TNFα, while both cytokines are capable of activating the p44/42 pathway. The greater than additive response is likely due to dual stimulation of the common p44/42 pathway. Finally, in order to simulate a physiological environment where obese, inflamed adipose tissue undergoes normal β-adrenergic stimulation during fasting and/or exercise, we treated adipocytes with a triple combination of isoproterenol, IL-6 and TNFα. This condition resulted in the highest level of lipolytic activation with a value slightly greater than adding isoproterenol stimulation to a combined treatment of IL-6 and TNFα (Figure 1. compare bar 8 with bars 1 and 7).\n\n\nConclusion\n\nThe level of lipolytic activation from the triple stimulation with isoproterenol, IL-6 and TNFα far surpassed that of normal β-adrenergic stimulation alone, and provides mechanistic evidence for the cause of hyperlipidemia in obese individuals. Under normal circumstances (in lean individuals), β-adrenergic signaling is activated during fasting and exercise to mobilize fatty acids from adipose tissue and compensate for a negative systemic energy balance. Once the energy balance has returned to homeostasis, β-adrenergic stimulation is inactivated and the release of fatty acids is halted to prevent excessive plasma lipid levels. Obese individuals are also subject to normal β-adrenergic, epinephrine and norepinephrine stimulation of adipose tissue due to stress responses or negative energy balance, and this stimulation signals through the heterotrimeric G-protein, adenylyl cyclase, cAMP, PKA network32. In addition to this signaling, cytokines produced in inflamed obese adipose tissue also activate additional pathways that make a cumulative addition to the normal lipolytic response. Evidence provided here suggests that IL-6- and TNFα-activated pathways in adipose contribute to increased lipolysis through both independent and common signaling pathways (Figure 2). In considering therapeutic options for obese individuals, maintaining normal β-adrenergic signaling is vital to manage routine changes in energy balance that occur due to cyclical variations in physical activity. However, the contributions of IL-6 and TNFα to increased lipolytic activity being additive to normal β-adrenergic stimulation indicates that these pathways (p44/42, JAK/STAT and JNK) represent excellent therapeutic targets that will prevent excessive lipolysis, yet minimally interfere with maintaining normal responses to varying energy demands.\n\nβ-adrenergic (epinephrine and norepinephrine) stimulation of adipose tissue to induce activation of lipolysis proceeds through heterotrimeric G-protein and adenylyl cyclase activation, which elevates cytosolic cAMP levels (green arrows). Cyclic-AMP-dependent protein kinase-A (PKA) then phosphorylates downstream regulatory proteins and lipases to initiate release of fatty acids from triacylglycerol stores. Inflamed obese adipose tissue is also subject to additional stimuli originating from secreted cytokines. In adipocytes, IL-6 activates the JAK/STAT and p44/42 pathways (blue/black arrows), while TNFα activates the JNK and p44/42 pathways (red/black arrows), which are independently sufficient to stimulate lipolysis. Based on the magnitude of fatty acid release when cytokine stimulation is concurrent with normal adrenergic signaling, it appears that the activation of combined pathways provides an additive response to lipolytic activation leading to an excess of fatty acids released.\n\n\nData availability\n\nF1000Research: Dataset 1. Glycerol release following isoproterenol, IL-6 and TNFα stimulation, 10.5256/f1000research.4151.d2771133", "appendix": "Author contributions\n\n\n\nNC and RO conceived the study. NC and RO designed the experiments. NC carried out the research. NC and RO prepared the first draft of the manuscript. NC, WG and RO contributed to the experimental interpretations and preparation of the manuscript. All authors were involved in revising 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 thank the University of New Mexico, School of Medicine, Office of Research for providing generous funding for this work.\n\n\nAcknowledgements\n\nWe thank Drs. Carolina Franco Nitta and Yijuan Sun for their thoughtful contributions to this study.\n\n\nReferences\n\nCalle MC, Fernandez ML: Inflammation and type 2 diabetes. Diabetes Metab. 2012; 38(3): 183–91. PubMed Abstract \| Publisher Full Text\n\nLolmede K, Duffaut C, Zakaroff-Girard A, et al.: Immune cells in adipose tissue: key players in metabolic disorders. Diabetes Metab. 2011; 37(4): 283–90. PubMed Abstract \| Publisher Full Text\n\nLukens JR, Dixit VD, Kanneganti TD: Inflammasome activation in obesity-related inflammatory diseases and autoimmunity. Discov Med. 2011; 12(62): 65–74. PubMed Abstract\n\nStohr R, Federici M: Insulin resistance and atherosclerosis: convergence between metabolic pathways and inflammatory nodes. Biochem J. 2013; 454(1): 1–11. PubMed Abstract \| Publisher Full Text\n\nLumeng CN, Bodzin JL, Saltiel AR: Obesity induces a phenotypic switch in adipose tissue macrophage polarization. J Clin Invest. 2007; 117(1): 175–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStrohacker K, McFarlin BK: Influence of obesity, physical inactivity, and weight cycling on chronic inflammation. Front Biosci (Elite Ed). 2010; 2: 98–104. PubMed Abstract\n\nGabay C: Interleukin-6 and chronic inflammation. Arthritis Res Ther. 2006; 8(Suppl 2): S3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStrieter RM, Kunkel SL, Bone RC: Role of tumor necrosis factor-alpha in disease states and inflammation. Crit Care Med. 1993; 21(10 Suppl): S447–63. PubMed Abstract\n\nWarren JS: Interleukins and tumor necrosis factor in inflammation. Crit Rev Clin Lab Sci. 1990; 28(1): 37–59. PubMed Abstract \| Publisher Full Text\n\nPlomgaard P, Fischer CP, Ibfelt T, et al.: Tumor necrosis factor-alpha modulates human in vivo lipolysis. J Clin Endocrinol Metab. 2008; 93(2): 543–9. PubMed Abstract \| Publisher Full Text\n\nvan Hall G, Steensberg A, Sacchetti M, et al.: Interleukin-6 stimulates lipolysis and fat oxidation in humans. J Clin Endocrinol Metab. 2003; 88(7): 3005–10. PubMed Abstract \| Publisher Full Text\n\nRyden M, Arner P: Tumour necrosis factor-alpha in human adipose tissue -- from signalling mechanisms to clinical implications. J Intern Med. 2007; 262(4): 431–8. PubMed Abstract \| Publisher Full Text\n\nChen X, Xun K, Chen L, et al.: TNF-alpha, a potent lipid metabolism regulator. Cell Biochem Funct. 2009; 27(7): 407–16. PubMed Abstract \| Publisher Full Text\n\nKlop B, Elte JW, Cabezas MC: Dyslipidemia in obesity: mechanisms and potential targets. Nutrients. 2013; 5(4): 1218–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMingrone G, DeGaetano A, Greco AV, et al.: Reversibility of insulin resistance in obese diabetic patients: role of plasma lipids. Diabetologia. 1997; 40(5): 599–605. PubMed Abstract \| Publisher Full Text\n\nArner P: Insulin resistance in type 2 diabetes: role of fatty acids. Diabetes Metab Res Rev. 2002; 18(Suppl 2): S5–9. PubMed Abstract \| Publisher Full Text\n\nGlund S, Krook A: Role of interleukin-6 signalling in glucose and lipid metabolism. Acta Physiol (Oxf). 2008; 192(1): 37–48. PubMed Abstract \| Publisher Full Text\n\nBastard JP, Jardel C, Delattre J, et al.: Evidence for a link between adipose tissue interleukin-6 content and serum C-reactive protein concentrations in obese subjects. Circulation. 1999; 99(16): 2221–2. PubMed Abstract\n\nFebbraio MA, Pedersen BK: Contraction-induced myokine production and release: is skeletal muscle an endocrine organ? Exerc Sport Sci Rev. 2005; 33(3): 114–9. PubMed Abstract \| Publisher Full Text\n\nSteensberg A, van Hall G, Osada T, et al.: Production of interleukin-6 in contracting human skeletal muscles can account for the exercise-induced increase in plasma interleukin-6. J Physiol. 2000; 529(Pt 1): 237–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-Khalili L, Bouzakri K, Glund S, et al.: Signaling specificity of interleukin-6 action on glucose and lipid metabolism in skeletal muscle. Mol Endocrinol. 2006; 20(12): 3364–75. PubMed Abstract \| Publisher Full Text\n\nCarey AL, Steinberg GR, Macaulay SL, et al.: Interleukin-6 increases insulin-stimulated glucose disposal in humans and glucose uptake and fatty acid oxidation in vitro via AMP-activated protein kinase. Diabetes. 2006; 55(10): 2688–97. PubMed Abstract \| Publisher Full Text\n\nLyngso D, Simonsen L, Bulow J: Metabolic effects of interleukin-6 in human splanchnic and adipose tissue. J Physiol. 2002; 543(Pt 1): 379–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWhite UA, Stephens JM: The gp130 receptor cytokine family: regulators of adipocyte development and function. Curr Pharm Des. 2011; 17(4): 340–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZvonic S, Baugh JE Jr, Arbour-Reily P, et al.: Cross-talk among gp130 cytokines in adipocytes. J Biol Chem. 2005; 280(40): 33856–63. PubMed Abstract \| Publisher Full Text\n\nHeinrich PC, Behrmann I, Haan S, et al.: Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochem J. 2003; 374(Pt 1): 1–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSethi JK, Xu H, Uysal KT, et al.: Characterisation of receptor-specific TNFalpha functions in adipocyte cell lines lacking type 1 and 2 TNF receptors. FEBS Lett. 2000; 469(1): 77–82. PubMed Abstract \| Publisher Full Text\n\nRyden M, Dicker A, van Harmelen V, et al.: Mapping of early signaling events in tumor necrosis factor-alpha -mediated lipolysis in human fat cells. J Biol Chem. 2002; 277(2): 1085–91. PubMed Abstract \| Publisher Full Text\n\nRyden M, Arvidsson E, Blomqvist L, et al.: Targets for TNF-alpha-induced lipolysis in human adipocytes. Biochem Biophys Res Commun. 2004; 318(1): 168–75. PubMed Abstract \| Publisher Full Text\n\nSouza SC, de Vargas LM, Yamamoto MT, et al.: Overexpression of perilipin A and B blocks the ability of tumor necrosis factor alpha to increase lipolysis in 3T3-L1 adipocytes. J Biol Chem. 1998; 273(38): 24665–9. PubMed Abstract\n\nLass A, Zimmermann R, Haemmerle G, et al.: Adipose triglyceride lipase-mediated lipolysis of cellular fat stores is activated by CGI-58 and defective in Chanarin-Dorfman Syndrome. Cell Metab. 2006; 3(5): 309–19. PubMed Abstract \| Publisher Full Text\n\nCollins S: β-Adrenoceptor Signaling Networks in Adipocytes for Recruiting Stored Fat and Energy Expenditure. Front Endocrinol (Lausanne). 2011; 2: 102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCard N, Garver WS, Orlando RA: Glycerol release following isoproterenol, IL-6 and TNFα stimulation. F1000Research. 2014. Data Source" }	[ { "id": "5232", "date": "09 Jul 2014", "name": "Marleen van Baak", "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 investigates the additive effects of isoproterenol, interleukin-6 and TNF-alpha on lipolysis in mature mouse 3T3-L1 adipocytes. For this, the maximal lipolytic doses of each of the three agonists were established and 3T3-L1 adipocytes were incubated for 24h with these doses, alone or in combination. Lipolysis was assessed as glycerol release into the medium over the 24h incubation period.The main conclusion to be drawn from the study is that IL-6 and TNF-alpha signaling are additive to the beta-adrenergic signaling with respect to 3T3-L1 adipocyte lipolysis. The investigators furthermore conclude \"that this suggests that therapeutic inhibition of cytokine signaling will prevent excessive lipolysis, yet minimally interfere with maintaining normal responses to varying energy demands\". This addition seems to be too far-fetched for several reasons: 1. confirmation in human adipocytes is required, because there may be species differences; 2. the study tested maximal lipolytic doses of the three lipolytic agents and under physiological conditions such a combination is unlikely to occur; 3. it is well-known that in obesity responsiveness to beta-adrenergic stimulation is blunted, therefore it could also be that the increased lipolysis due to increased cytokine levels in obesity helps to maintain normal lipolytic responsiveness to increased energy demand rather than result in excessive lipolysis.Other comments:Glucagon is mentioned several times as a stimulant of lipolysis without appropriate reference. There are few and mostly inconsistent data on a potential lipolysis-stimulating effect of glucagon in humans. Since the studies do not address glucagon, I would suggest not to include glucagon in the introduction and abstract. I find the term 'coincident' confusing. Although it is explained in the introduction what the authors mean by coincident, its use is not necessary. The question can simply be whether cytokines and isoproterenol are additive in their effects on lipolysis or not. In the Methods it is stated that the optimal dose to induce 'fatty acid hydrolysis' is identified. This should probably be 'triglyceride hydrolysis'. Figure 1 represents the results of three groups. In the Methods these three groups are not mentioned. In the legend to figure 1 it is mentioned twice that mean values from three replicate experiments are shown. Both in Figure 1 as well as in the methods there is reference to statistical testing. However, the results of these tests are not reported. The text in the Results section refers to bar numbers in Figure 1, but these are not indicated in the figure. Why do the authors conclude that the more than additive effect of Il-6 and TNF-alpha suggests that independent, as well as overlapping, signaling pathways were activated? Please explain more clearly.", "responses": [] }, { "id": "5475", "date": "21 Jul 2014", "name": "Guy Laureys", "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 evaluate the β-adrenergic pathwaywhich is an interesting target for modulation of inflammatory and metabolic pathways. This paper investigates the potential (additive) effects of β-adrenergic and cytokine stimulation on lipolysis in mouse adipose cells. Additive effects of isoproterenol, TNF-α and IL-6 on lipolysis in 3T3-L1 cells are demonstrated in vitro.A major question concerns the applied statistics:The n value seems insufficient to assume normal distribution, a non-parametric test seems more appropriate. The post-hoc test after ANOVA is not mentioned. If performed the (non)significance between different experimental groups should be indicated. Does the n=3 mean technical replicates? If so they should be averaged but not evaluated as separate experiments. If not, why does group 1 to 3 not cover all possible experimental strategies? This is confusing and should be clearly stated. Are there data on a sham experiment? If so they should be included.Other questions concerning this paper:As for most in-vitro studies: what is the relevance for the in-vivo and/or human situation? A good part of the discussion is based on assumptions - e.g. additive effects are suggested to be the consequence of the activation of different signaling pathways, however this is not experimentally verified.These questions should be addressed in an updated version of the paper.", "responses": [] } ]	1	https://f1000research.com/articles/3-134
https://f1000research.com/articles/3-113/v1	16 May 14	{ "type": "Method Article", "title": "The ‘SAR Matrix’ method and its extensions for applications in medicinal chemistry and chemogenomics", "authors": [ "Disha Gupta-Ostermann", "Jürgen Bajorath", "Disha Gupta-Ostermann" ], "abstract": "We describe the ‘Structure-Activity Relationship (SAR) Matrix’ (SARM) methodology that is based upon a special two-step application of the matched molecular pair (MMP) formalism. The SARM method has originally been designed for the extraction, organization, and visualization of compound series and associated SAR information from compound data sets. It has been further developed and adapted for other applications including compound design, activity prediction, library extension, and the navigation of multi-target activity spaces. The SARM approach and its extensions are presented here in context to introduce different types of applications and provide an example for the evolution of a computational methodology in pharmaceutical research.", "keywords": [ "Steadily growing numbers of active compounds provide a critically important knowledge base for medicinal chemistry but also challenge Structure-Activity Relationship (SAR) analysis1. For important therapeutic targets", "compound activity landscapes become increasingly complex2 and difficult to analyze. Increasing volumes and complexity of compound activity data require the development of computational approaches to effectively extract SAR information from heterogeneous sources1. In addition", "it is essential to make this information available in an intuitive form that can be appreciated in the practice of medicinal chemistry and utilized in compound design. Therefore", "a number of SAR visualization methods and graphical analysis tools have been developed in recent years2", "3 to view SAR characteristics of entire data sets or extract SAR information from compound activity data. Regardless of their algorithmic foundations and design specifics", "many (but not all) graphical analysis methods have in common that they provide a bird’s eye view of SAR information in compound data sets and depart from the single-series focus that has traditionally governed medicinal chemistry efforts. However", "multi-facetted SAR information obtained from heterogeneous compound sources must ultimately again be utilized to advance individual compound series", "which is a challenging task." ], "content": "Introduction\n\nSteadily growing numbers of active compounds provide a critically important knowledge base for medicinal chemistry but also challenge Structure-Activity Relationship (SAR) analysis1. For important therapeutic targets, compound activity landscapes become increasingly complex2 and difficult to analyze. Increasing volumes and complexity of compound activity data require the development of computational approaches to effectively extract SAR information from heterogeneous sources1. In addition, it is essential to make this information available in an intuitive form that can be appreciated in the practice of medicinal chemistry and utilized in compound design. Therefore, a number of SAR visualization methods and graphical analysis tools have been developed in recent years2,3 to view SAR characteristics of entire data sets or extract SAR information from compound activity data. Regardless of their algorithmic foundations and design specifics, many (but not all) graphical analysis methods have in common that they provide a bird’s eye view of SAR information in compound data sets and depart from the single-series focus that has traditionally governed medicinal chemistry efforts. However, multi-facetted SAR information obtained from heterogeneous compound sources must ultimately again be utilized to advance individual compound series, which is a challenging task.\n\nThe Structure-Activity Relationship Matrix (SARM) approach has originally been designed to extract and organize SAR-informative compound series from large data sets4 and has been further extended to help bridge the gap between data-driven SAR analysis, compound design, and activity predictions5 and study compound series in multi-target activity spaces6. Here, we present the SARM approach and its extensions in context and introduce new features and applications.\n\n\nMethods\n\nThe original design idea underlying the SARM approach was to systematically extract compound series with well-defined structural relationships from data sets and organize them in a matrix format4. To convey SAR information, matrix cells representing data set compounds are color-coded according to compound potency. The methodological basis for compound series identification and organization was provided by the matched molecular pair (MMP) concept7. An MMP is defined as a pair of compounds that differ only at a single site7. Compounds in MMPs can be interconverted by the exchange of a substructure, termed a chemical transformation8. In order to generate MMPs on a large scale, compounds must be systematically fragmented. The algorithm by Hussain and Rea8 (which we re-implemented and further modified in-house) provides an elegant and computationally efficient solution to this task by subjecting compounds to systematic deletion of individual exocyclic single bonds (single-cut) or simultaneous deletion of two (dual-cut) and three (triple-cut) exocyclic single bonds. The resulting fragments are then stored in an index table as keys (core structures) and smaller values (substituents)8.\n\nThe most important aspect of SARM design has been the application of dual fragmentation scheme leading to MMP generation at two levels4, as outlined in Figure 1. In the first step, MMPs are generated from data set compounds yielding “compound MMPs”. In the second step, core fragments from compound MMPs are again subjected to fragmentation leading to the generation of “core MMPs”. As a consequence, this hierarchical two-step fragmentation scheme identifies all compound subsets that have structurally analogous cores, i.e., core structures that are only distinguished by a structural modification at a single site. Each subset represents a so-called “structurally analogous matching molecular series” (A_MMS)4. Thus, each A_MMS represents a set of compound series with structurally analogous cores. Individual compounds and/or subsets of compounds can belong to multiple A_MMS, hence providing a high-level structural organization of a compound collection that captures all possible (MMP-based) substructure relationships.\n\nThree model series with three compounds each (A–C, D–F, and G–I) are shown with pKi values (red). In the first step, all compounds are fragmented at a single bond (purple dotted line) producing compound MMPs that yield a common core (key) and a compound specific substituents (values). In the second step, the cores resulting from the first step are further fragmented to obtain core MMPs. The SARM is then generated by combining series with structurally analogous cores that represent individual rows. In addition, colums represent substituents. In each cell, the combination of a core and a substituent defines a unique compound. Compounds present in the data set are indicated by filled cells that are color-coded according to potency using a continuous spectrum from red (low potency) over yellow to green (high). In addition, empty cells indicate virtual compounds. Substructures distinguishing the core fragments are highlighted in red.\n\nEach A_MMS is represented in an individual SARM, as illustrated in Figure 1. The SARM is filled with structurally analogous cores resulting from core MMPs (second fragmentation step) and the corresponding substituents obtained from compound MMPs (first fragmentation step). Single-, dual-, and triple-cut matrices are separately generated (vide supra). Each cell in a SARM represents a unique compound, i.e., a unique combination of a key and value fragment. Each row contains an individual analog series, i.e., compounds sharing the same core. Each column contains compounds from different series that share the same substituent (single-cut) or substituent combination (dual- or triple-cuts). The series forming a SARM typically contain different sets of substituents, giving rise to “real” compounds (filled cells) and “virtual” compounds (VC; empty cells). As also illustrated in Figure 1, a color spectrum is applied to represent the potency (or ligand efficiency) values of real compounds. Importantly, SARMs resemble standard R-group tables used in medicinal chemistry, although their design and information content is much more complex and comprehensive. Standard R-group tables typically only contain an individual core structure of a single series, all substituents, and associated potency values. However, because SARMs resemble R-group tables, they are readily accessible to medicinal chemists who can inspect individual compounds and their relationships to others.\n\nIn SARMs, different types of SAR patterns become readily apparent. This is illustrated in Figure 2 that shows exemplary SARMs revealing characteristic patterns (for representation purposes, only small matrices are shown; vide infra). For example, the SARM in Figure 2a identifies two preferred core structures that consistently produce potent compounds. Furthermore, the SARM in Figure 2b reveals an SAR transfer event, i.e., the presence of two compound series with related yet distinct core structures that contain pairwise corresponding analogs with similar potency progression. Other SAR patterns that can frequently be detected include, for example, preferred R-groups (or R-group combinations) in related compound series or regions of distinct SAR continuity or discontinuity. Continuous SAR regions are characterized by the presence of compounds with structural modifications that lead to gradual changes in potency, whereas discontinuous SAR regions contain structural analogs with large (and essentially unpredictable) potency variations2.\n\nIn (a), a SARM capturing 16 ligands of the histamine H4 receptor is shown. Cells containing analogs with preferred cores yielding potent compounds are framed in blue. Substructures distinguishing the core fragments are highlighted in red. The pKi value range for the 16 ligands is displayed. In (b), a subset of a double-cut SARM is shown that contains series of carbonic anhydrase I inhibitors and an exemplary SAR transfer event.\n\nLarge compound data sets typically yield many SARMs of different size and composition, depending on their degree of structural homogeneity or heterogeneity. Two examples are given to illustrate this point. First, an in-house focused compound library with various substitutions of a small number of core structures comprising 6503 compounds produced a total of 6738 (single-, double- and triple-cut) matrices containing a total of 135,619 VCs. Second, a structurally heterogeneous set of 509 purinergic receptor (P2Y12) ligands generated at total of 181 SARMs containing 17,445 VCs. Again, each SARM contains a unique A_MMS and individual compounds might belong to multiple A_MMS depending on the structural relationships they form. SARMs provide highly resolved views of all of these structural relationships. Depending on the number of compounds forming A_MMS, the size of SARMs can considerably vary. For example, in a survey of 32 different activity classes consisting of 398 to 2497 compounds, SARMs were found to contain between three and 555 compounds, with a median value of 13. Furthermore, we also use a “matrix overlap” measure to account for the ratio of data set compounds versus VCs, which typically varies in SARMs. Matrix overlap is determined as the average over all row overlap values. For individual columns in SARMs, row overlap (RO) is calculated as:\n\n\n\nwhere, ‘real compounds’ correspond to the number of data set compounds present in each column. RO yields a numerical score between 0 (no overlap) and 1 (complete overlap). Figure 3 reports the matrix overlap distribution for SARMs from the focused library referred to above, which is a fairly representative distribution for structurally homogeneous data sets. As an alternative measure, “matrix coverage” (C), which accounts for the proportion of cells in a SARM that are populated with real compounds n can be calculated as:\n\n\n\nShown is a histogram with the matrix overlap distribution for SARMs from an in-house focused library.\n\nRegardless of the number of SARMs that are obtained from large data sets, there are too many for one-by-one inspection. Hence, ranking schemes should be applied to prioritize and pre-select those SARMs that are most informative for a given application. For instance, SARMs can be easily ranked on the basis of numerical functions that prioritize matrices containing preferred substituent combinations or core structures and SAR transfer events or matrices that capture high degrees of local SAR continuity or discontinuity. For example, Figure 4 shows two SARMs originating from a large data set that are highly ranked on the basis of SAR discontinuity (as indicated by the presence of multiple analogs with large potency differences). Depending on the applied selection criteria, most informative SARMs can be readily inspected on the basis of a ranked list.\n\nIn (a) and (b), two SARMs are shown (resulting from single- and triple-cut fragmentation, respectively) for corticotropin-releasing factor receptor 1 ligands that were highly ranked on the basis of SAR discontinuity scoring.\n\nVCs contained in SARMs provide immediate suggestions for compound design. Because VCs represent unexplored key-value combinations derived from data set compounds, the union of VCs from all SARMs provides a “chemical space envelope” for a given compound set or library. VCs originating from SAR-informative matrices represent natural focal points for interactive compound design. Moreover, the potency of many virtual compounds can be predicted by applying a compound neighborhood (NBH) principle5, as illustrated in Figure 5. An NBH of a given VC is defined by three adjacent real compounds that contain the core of the VC (compound G in Figure 5), its substituent (compound E) and the core and substituent of G and E (compound D). The potency of the VC can then be predicted by applying the additivity assumption underlying Free-Wilson analysis9 using the simple equation shown in Figure 5. The putative potency value of the VC results from the sum of (logarithmic) potencies of the two real compounds sharing the same core and substituent with the VC, respectively, minus the potency of the compound that contains the core structure and substituent of the two other real compounds. Thus, from NBHs, “mini-QSAR” models are derived for activity prediction. For each candidate VC, qualifying NBHs are collected across all SARMs, individual potency predictions are carried out, and their consistency is evaluated, for example, by calculating standard deviations for predictions5. In benchmark calculations on six different sets of G protein-coupled receptor ligands, potency values of subsets of test compounds falling into continuous local SAR regions were accurately predicted using the NBH-based approach, and prediction accuracy generally increased with the number of qualifying NBHs5. This is also relevant for practical applications. For potency prediction, candidate VCs should be prioritized for which multiple NBHs are available. For example, for the set of 509 purinergic receptor ligands (vide supra), 5167 of 17,445 VCs were found to have at least three qualifying NBHs. Hence, in these cases, the consistency of potency predictions can be assessed. Such candidate VCs can be explored in a systematic manner. For libraries tested in individual assays, VCs predicted to be consistently active on the basis of multiple NBHs provide preferred candidates for target/assay-dependent library expansion and focusing.\n\nAn NBH of virtual compound X is marked in blue in a model SARM and displayed in detail. Compounds E and G share the same substituents and core with X, respectively, and the third neighbor D combines the core and substituent of E and G, respectively. At the bottom, the equation to predict the potency of X from the potency values of E, G, and D is shown.\n\nImportantly, the NBH-based mini-QSAR approach is only applicable to candidate compounds falling into SARMs that represent continuous SAR regions, as illustrated in Figure 6a. By contrast, compounds falling into discontinuous SAR regions, as shown in Figure 6b, fall outside the applicability of standard QSAR modeling. Nonetheless, VCs from SARMs representing discontinuous SAR regions are also attractive candidates for compound design. This especially applies to VCs falling into the vicinity of activity cliffs10 that are formed by pairs of structural analogs with large potency differences, as illustrated in Figure 6b. Activity cliffs represent the pinnacle of SAR discontinuity. VCs in the vicinity of activity cliff can often be expected to display large (positive or negative) potency fluctuations and are hence attractive candidates in the search for potent hits. Although a QSAR formalism cannot be applied to predict the potency of such compounds, they can be easily selected from SARMs containing activity cliffs on the basis of a “guilt-by-association” principle, i.e., VCs are preferentially selected that are neighbors of potent activity cliff partners. For this purpose, SARMs capturing high degrees of local SAR discontinuity are selected on the basis of discontinuity ranking (vide supra).\n\nIn (a), a SARM is shown that represents a highly continuous local SAR environment. In this case, the potency of a virtual compound can be predicted using the NBH-based approach. By contrast, (b) shows a SARM representing a discontinuous local SAR. Activity cliff-forming compound pairs are highlighted in blue. Such SAR environments fall outside the applicability domain of NBH-based potency predictions. However, marked VCs represent promising candidates for compound design based on their proximity to activity cliffs. Both SARMs originate from a set of cannabinoid CB1 receptor ligands (compound structures are omitted for clarity).\n\nSARMs have also been adapted for the navigation of multi-target activity spaces, which are populated by promiscuous compounds. In this context, promiscuity is defined as the ability of a compound to specifically interact with multiple targets (as opposed to non-specific binding effects)11. Here, the primary purpose of the matrix approach is not SAR analysis, but the systematic exploration of compound promiscuity patterns. Therefore, matrices capturing multi-target activities are generated. Such matrices have been designated as Compound Series Matrices (CSMs)6. CSMs are of interest for chemogenomics applications in which compound-target interactions are systematically explored12. In Figure 7, two exemplary CSMs of different composition and target coverage are shown that reveal different compound promiscuity patterns. In CSMs, data set compounds are color-coded according to the number of targets they are active against (instead of potency-based coloring). In Figure 7a, two structural analogs display very different degrees of promiscuity and in Figure 7b, a center of promiscuity is identified in a sparsely populated matrix. CSMs are designed to mine chemogenomics data sets and also offer immediate suggestions for the design of compounds with different multi-target activities. In addition, it is also readily possible to deconvolute CSMs into individual single-target SARMs, as illustrated in Figure 8. This makes it possible to compare SARMs across different targets and identify compounds that are attractive candidates for testing against additional targets.\n\n(a) shows a CSM containing 15 inhibitors of eight carbonic anhydrase (CAR) isoforms. Target coverage of analogs is reflected by increasingly dark blue shading of cells. Substructures distinguishing the core fragments are highlighted in red. The matrix composition is summarized (top left) and the target profile reported (top right). (b) shows a CSM with 44 analogs active against nine targets (including the HERG anti-target) belonging to three different families. The maximum common core structure (MCS) of the analog series is displayed. For clarity, compound structures are omitted. Target abbreviations: 5-HT; serotonin receptor, ST; serotonin transporter, DOP; dopamine receptor, HERG; HERG ion channel.\n\nThe deconvolution of a CSM with eight analogs active against the dopamine D2, D3, and D4 receptor isoforms into three single-target SARMs is illustrated. In all matrices, cells corresponding to VCs are not color-coded. In SARMs, cells of compounds with no available activity annotation for a given target are colored gray.\n\nJava programs were written, in part with the aid of the OpenEye chemistry tool kit13, to identify A_MMS and generate, rank, and display SARMs. Routines for potency predictions were also implemented in Java. Statistical analyses were carried out using R14. All compounds shown herein were obtained from ChEMBL15.\n\n\nConcluding remarks\n\nHerein, we have reviewed the design of the SARM methodology and discussed recent extensions and selected applications. In-house implementations of the SARM approach have been continuously developed and further refined to increase the utility of the methodology for medicinal chemistry. Primary reasons for discussing the different aspects and applications of SARMs in context have been to expose this approach to a wider drug development audience and provide an example for the data- and application-driven evolution of a computational medicinal chemistry method. SARMs can essentially be rationalized as local activity landscapes of data sets that are based upon a unique and comprehensive structural organization. SARMs primarily focus on activity information associated with series of closely related compounds but can also be applied to systematically study compound promiscuity patterns. In addition, they can also be easily adapted to explore other structure-property relationships relevant to drug discovery. A special feature of SARMs that sets them apart from many other activity landscape representations is that they closely link descriptive compound data analysis (a primary task of activity landscape modeling) and prospective compound design. Because SARMs are reminiscent of conventional R-group tables, they are readily accessible to medicinal chemists, thus circumventing the communication barrier that often hinders the effective application of computational approaches in the practice of medicinal chemistry. Future research activities will focus on the design of multi-property SARMs to aid in advanced compound optimization efforts.", "appendix": "Author contributions\n\n\n\nJB conceived the study, DGO collected the data and generated the representations, JB wrote the manuscript, and both authors examined 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\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors thank Dr. Anne Mai Wassermann, Dr. Dilyana Dimova, and Dr. Preeti Iyer for key contributions to SARM method development and applications.\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\nWassermann AM, Wawer M, Bajorath J: Activity landscape representations for structure-activity relationship analysis. J Med Chem. 2010; 53(23): 8209–8223. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Bajorath J: Methods for SAR visualization. RSC Adv. 2012; 2(2): 369–378. Publisher Full Text\n\nWassermann AM, Haebel P, Weskamp N, et al.: SAR matrices: automated extraction of information-rich SAR tables from large compound data sets. J Chem Inf Model. 2012; 52(7): 1769–1776. PubMed Abstract \| Publisher Full Text\n\nGupta-Ostermann D, Shanmugasundaram V, Bajorath J: Neighborhood-based prediction of novel active compounds from SAR matrices. J Chem Inf Model. 2014; 54(3): 801–809. PubMed Abstract \| Publisher Full Text\n\nGupta-Ostermann D, Hu Y, Bajorath J: Systematic mining of analog series with related core structures in multi-target activity space. J Comput Aided Mol Des. 2013; 27(8): 665–674. PubMed Abstract \| Publisher Full Text\n\nKenny PW, Sadowski J: Structure modification in chemical databases. In Chemoinformatics in Drug Discovery; Oprea, T. I., Ed.; Wiley-VCH: Weinheim, Germany, 2005; pp 271–285. Publisher Full Text\n\nHussain J, Rea C: Computationally efficient algorithm to identify matched molecular pairs (MMPs) in large data sets. J Chem Inf Model. 2010; 50(3): 339–348. PubMed Abstract \| Publisher Full Text\n\nKubinyi H: Free Wilson analysis. theory, applications and its relationships to Hansch analysis. Quant Struct Act Relat. 1988; 7(3): 121–133. Publisher Full Text\n\nStumpfe D, Bajorath J: Exploring activity cliffs in medicinal chemistry. J Med Chem. 2012; 55(7): 2932–2942. 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\nBajorath J: Computational approaches in chemogenomics and chemical biology: current and future impact on drug discovery. Expert Opin Drug Discov. 2008; 3(12): 1371–1376. PubMed Abstract \| Publisher Full Text\n\nOEChem, version 1.7.7, OpenEye Scientific Software, Inc., Santa Fe, NM, USA, 2012. Reference Source\n\nR: A Language and environment for statistical computing; R Foundation for statistical computing, Vienna, Austria, 2008. Reference Source\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" }	[ { "id": "4803", "date": "10 Jun 2014", "name": "Herman van Vlijmen", "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 on the SAR matrix method offers a useful approach to extract the relevant information from large datasets with compound activity data, and to present this in an intuitive way to chemists and computational chemists. It is therefore an attractive tool to use in analysis of HTS screens and to find structure activity trends and discontinuities in large groups of structurally related molecules.The title and abstract cover the content well. The chemogenomics application of the method is related to the use of compound promiscuity instead of the more usual compound activity on a single target. The methods are clearly described and can most likely be reproduced. The datasets used (even the dataset from the public source CHEMBL) are not provided, so the results will be difficult to reproduce. There is no mention in the manuscript on the availability of the tools that were developed. Methods like these could get widespread usage if they would be available to a wider audience. It would also be good if at least the public dataset would be made available so the results can be compared to other approaches.One significant benefit of this approach is that a large dataset can automatically be processed by the method to create multiple (often very many) SAR matrices. The authors point out that the idea is not that all these (often hundreds or thousands) SAR matrices are inspected visually, but that interesting elements in the matrices can be identified automatically, for instance \"virtual compounds\" (core-substituent combinations not yet made) that are predicted to have interesting activity. The same virtual compound can appear in different SAR matrices, and therefore multiple predictions can be made for the same compound and the level of consistency could be a good indicator for deciding to make he actual compound. The method also automatically identifies virtual compounds that are close to activity cliffs and are therefore interesting to make and test.", "responses": [] }, { "id": "4991", "date": "11 Jun 2014", "name": "Georgia B. McGaughey", "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 article with a sound description of the Structure-Activity Relationship (SAR) Matrix (SARM) methodology. Only a few questions / suggestions are recommended.In the Methods section, you describe the definition of a MMP. Specifically, you rely on the algorithm by Hussain and Rea. Have you considered MMPs where the only change is in a ring (i.e., an aromatic versus partially saturated ring)? Continuing onwards in the Methods section, you extend the MMP concept to include Matched Molecular Series (MMS). While I believe you were the first to coin this description, there are now others who are also using this formalism (e.g. NextMove (Roger Sayle)) and some reference to these additional methods is advised, particularly to avoid confusion since the names are similar. If there are differences, perhaps you could expand on them. In the Matrix distribution and ranking section, could you expand on the matrix overlap, row overlap and matrix coverage with at least one specific example (ie show the math in the supplementary material)? Additionally, I recommend using the same variables (i.e. are \"n\" and \"#real compounds\" the same)? If so, then they should be consolidated to one variable. Additionally, it's not obvious how the numbers to the right (5%) and left (30%) of the two matrices are derived. There appears to be some mis-counting of the number of targets in both Figure 7a & b. For Figure 7a, are there 10? And for Figure 7b, we count 10 targets and 4 families. Make sure to check the Figure 7 caption. Consider referring to DOP-D2 as merely D2; likewise DOP-D3 should be merely D3. Check the commas and semicolons as there are inconsistencies. Can these multi-target compound series matrices be tied to Adverse Drug Reactions (ADR)? That seems as a possible extension of this current work. In Figure 7b you consider hERG and hence, it seems as an opportunity to extend beyond merely the primary activity. Consider changing the word \"accessible\" in the concluding remarks to \"intuitive\" or \"interpretable\".", "responses": [] } ]	1	https://f1000research.com/articles/3-113
https://f1000research.com/articles/3-67/v1	07 Mar 14	{ "type": "Research Article", "title": "Evidence from molecular dynamics simulations of conformational preorganization in the ribonuclease H active site", "authors": [ "Kate A. Stafford", "Arthur G. Palmer III", "Kate A. Stafford" ], "abstract": "Ribonuclease H1 (RNase H) enzymes are well-conserved endonucleases that are present in all domains of life and are particularly important in the life cycle of retroviruses as domains within reverse transcriptase. Despite extensive study, especially of the E. coli homolog, the interaction of the highly negatively charged active site with catalytically required magnesium ions remains poorly understood. In this work, we describe molecular dynamics simulations of the E. coli homolog in complex with magnesium ions, as well as simulations of other homologs in their apo states. Collectively, these results suggest that the active site is highly rigid in the apo state of all homologs studied and is conformationally preorganized to favor the binding of a magnesium ion. Notably, representatives of bacterial, eukaryotic, and retroviral RNases H all exhibit similar active-site rigidity, suggesting that this dynamic feature is only subtly modulated by amino acid sequence and is primarily imposed by the distinctive RNase H protein fold.", "keywords": [ "Ribonuclease H1 (RNase H) proteins are well-conserved endonucleases that are found in all domains of life and cleave the RNA strand of an RNA-DNA duplex substrate. The RNase H active site canonically consists of a highly conserved DED(D) motif (Figure 1)", "three to four carboxylate-containing residues collectively participating in the binding of catalytically required divalent cations", "Mg2+ under physiological conditions. This active-site sequence motif and requirement for Mg2+ is widely shared with other nucleases", "suggesting a common catalytic mechanism1." ], "content": "Introduction\n\nRibonuclease H1 (RNase H) proteins are well-conserved endonucleases that are found in all domains of life and cleave the RNA strand of an RNA-DNA duplex substrate. The RNase H active site canonically consists of a highly conserved DED(D) motif (Figure 1), three to four carboxylate-containing residues collectively participating in the binding of catalytically required divalent cations, Mg2+ under physiological conditions. This active-site sequence motif and requirement for Mg2+ is widely shared with other nucleases, suggesting a common catalytic mechanism1.\n\n(A) Residue conservation among bacterial RNase H homologs mapped onto the structure of ecRNH (PDB ID 2RN2). Highly conserved residues are shown in red, highly variable residues in green, and sites with insufficient data in yellow. Image produced using ConSurf. (B) Electrostatic map of the solvent-accessible surface of ecRNH produced using APBS. Red represents regions of negative charge and blue represents regions of positive charge. The active-site residues are represented as sticks in both cases.\n\nThe best-studied member of the RNase H family is the homolog from Escherichia coli (ecRNH), in which this active-site motif is represented as D10, E48, D70, and D1342 (Figure 2A). Measurements of the pKa values of the active-site residues indicate perturbed pKa values for D10 and D70 which normalize upon Mg2+ binding, clearly establishing these residues as critical for interaction with ions3. The pH optimum for the RNase H reaction in vitro is approximately 7.5–8.54, a value at which all active-site residues should be deprotonated3 and therefore accessible for ion binding.\n\nIn all cases the backbone and active-site sidechains from ecRNH in the absence of ion (PDB ID 2RN2) are shown in light blue for comparison. (A) Structural superposition of the four active-site residues in two structures of ecRNH in the absence of metal ions: 2RN2 (light blue) and 1RNH (dark blue). (B) Structural diversity of RNases H in complex with a single Mg2+ ion: ecRNH (1RDD), green; XMRV WT (4E89), dark cyan; XMRV ∆C (3P1G), maroon; MoMLV ∆C (2HB5), purple. The two deletion mutants (indicated as ∆C) bind Mg2+ in slightly different positions than do their corresponding full-length proteins and both contain two alternate conformations for E48 and D134. (C) Comparison of Mg2+ and Mn2+ complexes: ecRNH with Mg2+ (1RDD), green; ecRNH with Mn2+ (1G15), orange; HIV RNase H domain with ecRNH helix C insertion with Mn2+ (3HYF), brown. (D) Structural diversity of RNases H in complex with substrate: Homo sapiens RNase H with Ca2+ ions (2QKK), brown; Bacillus halodurans RNase H with Mn2+ ions (1ZBI), dark green.\n\nDespite extensive study, the interaction of metal ions with the ecRNH active site is poorly understood. Activity has been reported in the presence of Mn2+ as well as the physiologically relevant Mg2+5. Significant differences have been observed between the protein’s interactions with Mg2+ and Mn2+. Co-crystallization studies of ecRNH with high concentrations of Mg2+ find a single bound metal ion6 (Figure 2B). By contrast, co-crystallization with Mn2+ reveals two bound ions, one associated with residues D10 and D134 (denoted the A site), and one associated with D10, E48, and D70 (denoted the B site)7 (Figure 2C); the B site is similar but not identical to the previously identified Mg2+ site. Single Mn2+ sites have been identified in the presence of mutations of E48 and/or D1348, both of which are dispensable for Mn2+-dependent activity9. Crystallographic studies of related RNases H from the archaeal extremophile Bacillus halodurans10 and from Homo sapiens11 in complex with substrate find two bound ions in the active site (Figure 2D).\n\nExperimental evidence from nuclear magnetic resonance (NMR) studies locates the area surrounding the active site as the region most susceptible to perturbation upon interaction with ions (Figure 3). Titration of Mg2+ with ecRNH, monitored independently by 1H and 25Mg2+ NMR, suggests that only a single ion binds to the protein in the absence of substrate5. The identified binding site has relatively weak affinity; Kd has been reported in the micromolar5 to low millimolar range12. The second site may be occupied only upon binding of substrate8, possibly due to the presence of high local concentration in the ion cloud of the highly negatively charged nucleic acid molecule. Conformational changes in the active site upon binding the first ion have been suggested as well, with the second site proposed as being responsible for the attenuation of activity at high ion concentrations13. Collectively, these results have been used to propose both a one-metal7,8,14 and a two-metal10,15,16 catalytic mechanism. Computational work using the quantum mechanics/molecular mechanics (QM/MM) method applied to the Bacillus halodurans17–19 and Homo sapiens20 complexes generally supports the two-metal mechanism.\n\n(A) Active site residues (yellow) and other DENQ residues (green) that experience perturbation upon Mg2+ binding24. (B) Chemical shift perturbation values for sidechain Cγδ reflecting the effects of Mg2+ binding24. White corresponds to no chemical shift change, red corresponds to a large change, and non-DENQ residues are shown in light blue. (C) Residues previously shown to experience backbone 15N or 1H chemical shift changes upon binding Mg2+25.\n\nRNase H domains are widely distributed in cellular organisms, but also occur as a component of the reverse transcriptase protein found in retroviruses, in which they are required for viral proliferation21. For this reason, inhibitors of retroviral RNase H domains, particularly that of HIV, have been widely reported15,22, although none to date have reached clinical use. Most such inhibitors interact with the active site in the metal-bound state22 and therefore must be selective for retroviral RNase H domains to find clinical utility. The HIV RNase H domain (hivRNH) has been reported to bind two metal ions even in the absence of substrate23, although the reason for this difference in behavior between hivRNH and ecRNH is not clear. The Homo sapiens RNase H domain (hsRNH) has not been structurally characterized in the absence of substrate and its binding behavior is less well understood. However, it has higher sequence identity and is more structurally similar to ecRNH than hivRNH.\n\nCombined nuclear magnetic resonance (NMR) and molecular dynamics (MD) studies of the behavior of the ecRNH active site residues suggest that the residues of the ecRNH active site are preorganized in the apo state for the binding of a single Mg2+ ion24. However, experimental constraints prevent the detailed observation of the protein’s dynamic behavior in the presence of a bound ion at ps-ns timescale. The present work aims to more fully understand the dynamics of ecRNH in the Mg2+-bound state through molecular dynamics simulations of ecRNH in the presence of single Mg2+ ions in various positions in the active site as suggested by crystallographic studies. In addition, the dynamic behavior of the active site in the apo state is compared with homologs from other organisms.\n\n\nMethods\n\nFor each initial protein structure, protonation states for titratable residues were assigned either by experimental measurement (for ecRNH3) or by prediction using the H++26 pKa predictor. Unless otherwise specified, all simulations were performed at a pH of 5.5 to recapitulate the conditions used in prior NMR experiments on ecRNH27,28. Crystallographic water molecules were removed from all structures prior to solvation using Schrodinger’s Maestro tool, version 8.5 or 9.1, as distributed in the Desmond software package. Simulations were performed as described24,29 using Desmond academic release 3 or source release 2.4.2.130. Proteins were described with the Amber99SB force field31, solvated with TIP3P water32 in a cubic box with a 10Å buffer region from solute to box boundary, and neutralized with Cl− ions. Bonds to hydrogen atoms were constrained using the M-SHAKE algorithm33. Simulations containing Mg2+ ions used the Aqvist parameter set34. Electrostatics were calculated with the PME method using a 9Å cutoff. All simulations used a 2.5fs inner timestep on a 1-1-3 RESPA cycle and were performed in the NVT ensemble using a Nosé-Hoover thermostat after equilibration to constant box volume for 5ns in the NPT ensemble. All simulations described in this work were run for 100ns unless otherwise noted. These simulation conditions applied to the apo state of RNase H homologs have previously been shown to reproduce NMR data well35.\n\nOrder parameters were calculated by the equation36:\n\n\n\nin which µi and µj represent the x, y, and z components of a unit vector μ→ in the direction of a given chemical bond. This represents the long-time limit of the angular reorientational correlation function for a given bond vector.\n\nProtein Data Bank (PDB, RRID:nif-0000-00135) structures used for initiating trajectories are listed, along with their resolutions and any system-specific preparation steps, in Table 1.\n\nSummary of crystal structures used as starting points for molecular dynamics simulations of RNase H homologs.\n\n\nResults and discussion\n\nA simulation was initiated from the crystal structure of ecRNH in the Mg2+-bound state (PDB ID 1RDD)6. However, the position of the ion identified in this structure is not stable in simulation and exits the binding site immediately upon initiation of the trajectory. The ion transiently interacts with the protein at a variety of sites on the protein surface over the course of the 89ns trajectory but never returns to its original position in the active site (Figure 4A).\n\n(A) Occupancy map from an 89ns simulation initiated from the ecRNH structure solved in the presence of Mg2+ (PDB ID 1RDD), contoured to 0.05% occupancy (corresponding to at least 45ps total residence time). The ion exits the active site and interacts with a variety of regions on the protein surface. (B) The active-site region of the 1RDD structure, colored by atomic B-factor. The B-factor of the ion is substantially larger than the surrounding residues, and is in fact larger than the B-factor of any other atom in the structure save crystallographic waters.\n\nHistorically, simulation of the behavior of multivalent ions using standard molecular mechanics force fields has been a long-standing challenge44. It is therefore possible that the instability of this position in simulation is an artifact of force field errors. However, given that ions in this position are not observed in the substrate-bound structures of RNase H homologs (Figure 2D), and that the B-factor of the Mg2+ ion in the 1RDD structure is much higher than those of the surrounding residues (Figure 4B), it is likely that this position does not reflect the most stable conformation of the protein-ion complex in solution.\n\nAdditional simulations were carried out under the same conditions for single Mg2+ ions in each of the two Mn2+ binding sites identified for ecRNH. Because the crystal structure of ecRNH solved in the presence of Mn2+ (PDB ID 1G15) exhibits disorder in both the active-site and handle loops7, the ion positions were instead modeled into the apo ecRNH structure (PDB ID 2RN2) by superposition. For the model of the B-site Mg2+ ion, the rotamer of E48 was also corrected to match the orientation observed in the 1G15 structure. For comparison to an alternative homolog, the Mg2+ ion in the B site was also modeled into the RNase H structure from the thermophilic bacterium Thermus thermophilus (ttRNH, PDB ID 1RIL), whose structure was also solved in the absence of divalent ions38.\n\nMg2+ ions were found to be stably associated with the ecRNH active site in both simulations, despite the fact that the ions were modeled into a structure that did not originally contain them (Figure 5). This observation clearly supports the hypothesis that conformational preorganization in the active site promotes ion binding. It is possible that the effectiveness of this modeling procedure was facilitated by a well-documented feature of crystal packing in ecRNH, in which the amino group of a lysine sidechain in a neighboring molecule inserts into the negatively charged active site in a position approximating the B site8. However, a short simulation of ttRNH, whose structure does not contain this contact, with Mg2+ modeled into the B site was also stable, suggesting that crystal contacts in ecRNH are not responsible for the observation of preorganization in its active site.\n\n(A) Occupancy maps contoured at 10% occupancy for the A site (blue) and B site (green). Neither ion leaves the active site over the 100ns trajectory length. (B) Closer view of the active site in which the four active-site residues in apo ecRNH and the positions of the two Mn2+ from which the trajectories were initiated (derived from PDB ID 1G15) are shown for comparison.\n\nThe presence of Mg2+ located in either the A or the B site did not substantially affect the dynamics of the active-site residues as determined by S2. All four residues remain highly rigid in the presence of a Mg2+ ion in either position (Figure 6). The major difference between the unbound, A site, and B site trajectories’ sidechain dynamics was observed in a short loop between helix D and β-sheet 5. Experiments have demonstrated dynamics in this region on the ps-ns timescale, suggesting that the loop is simply incompletely sampled in 100ns simulations rather than significantly perturbed by ion binding. No significant differences in the behavior of these residues upon introduction of ions are observed experimentally24.\n\nCalculated S2 values are shown for various simulation conditions with standard errors of the mean: ecRNH apo (light blue), A site (dark blue), B site (green). The four active-site residues are indicated with red triangles. In no case does the difference between any two simulations reach statistical significance.\n\nOf the four conserved catalytic residues, D134 is known to be somewhat dispensable; catalytic activity is retained, though reduced, by substitutions with N or H, which also increase thermostability45. In conjunction with crystallographic evidence, this suggests that the B site is occupied in the absence of substrate. Because measurements of the sidechain 13Cγδ resonances by NMR could not clearly distinguish the behavior of D134 (the unique participant in the A site) from E48 (the unique participant in the B site)24, comparisons of the two trajectories provide an additional opportunity to distinguish between these two sites.\n\nAlthough single metal ions in both sites were stably bound to the protein, the RMSD over the course of each 100ns trajectory was larger for the ion in the A site (1.2Å) compared to the B site (0.6Å), which in turn is similar to the RMSD of a 30ns control simulation of ttRNH with an ion modeled into the B site (0.6Å). Additionally, a small amount of motion in the direction of the B site was observed for the ion in the A site; the initial and final positions differ by 1.7Å (Figure 5). (By comparison, the A and B sites are about 4Å apart.)\n\nDistinct conformations were also observed for several neighboring residues, reflecting reorganization of local hydrogen bonding networks to accommodate ion binding in each of the two sites. N45 does not differ significantly in sidechain rigidity between the two trajectories, but it does differ in conformation: in the A site trajectory, it is oriented away from the substrate-binding site and participates in a network of interactions that also includes the conserved site T43, while in the B site trajectory N45 is primarily oriented into solvent and occupies the rotamer found in the hsRNH-substrate complex.\n\nThe hydrogen-bonding network surrounding D134 unsurprisingly differs considerably between the A and B site trajectories. Occupancy of inter-sidechain hydrogen bonds in this region is summarized in Table 2. H124, which interacts with substrate in the hsRNH complex and is known to be associated with product release, forms hydrogen bonds with D134 in the B site trajectory, partially displacing one of the hydrogen bonds formed between D134 and R138 in the apo trajectory. By contrast, H124 interacts primarily with E131 in the A site trajectory, while D134 coordinates Mg2+ in a monodentate manner, partially displacing the R138-D134 interaction. This conformation too is at odds with experimental evidence, since E131 experiences minimal chemical shift perturbation upon Mg2+ binding24. Examination of the hsRNH-substrate complex reveals that R138 participates in a hydrogen-bonding network that includes D134 and the phosphate adjacent to the scissile phosphate (Figure 7); hydrogen-bonding interactions in the apo state may thus minimize entropic costs of binding.\n\nInteractions between D134, R138, metal ion A, and the phosphate backbone of the RNA strand are shown (PDB ID 2QKK). An asterisk indicates the scissile phosphate. The preorganization of the R138-D134 salt bridge in the apo state of ecRNH likely minimizes the entropic cost of forming this interaction upon substrate binding.\n\nComparison of hydrogen-bonding environments in the network surrounding the active-site residue D134 in the apo trajectory of ecRNH compared to trajectories containing an Mg2+ ion in either the A or the B site. Hydrogen bonds were considered formed if the donor-acceptor distance was less than 3.1Å and the donor-hydrogen-acceptor angle was less than 25°.\n\nThese results collectively add to prior experimental evidence that the B site is the primary site for metal ion binding in the absence of substrate. Furthermore, the presence of a metal ion in the B site may induce reorganization of the surrounding sidechains into conformations conducive to subsequent substrate binding.\n\nGiven that all known RNase H homologs have extremely similar active-site structures, it is likely that measurements made on the ecRNH protein can be generalized to other RNase H homologs. S2MD values were therefore calculated from previous simulations of the four handle-region-containing bacterial RNase H homologs of known structure, as well as for hsRNH in the absence of substrate29.\n\nAs might be expected from the high level of structural conservation in the active-site region, the five handle-region-containing RNase H homologs compared differ very little in the dynamics of their active site residues (Figure 8). Notably, the trajectory initiated from the hsRNH structure, which was solved in the presence of substrate and which contained a Na+ ion in a position similar to the B site in ecRNH, differs very little from trajectories initiated from any other RNase H structure lacking these additional components. This observation provides strong support for the interpretation that the rigid active-site residues are conformationally preorganized for metal-ion interactions even in the unbound state.\n\nCalculated S2 values are shown for the four active-site residues in RNase H homologs. Left: soRNH (dark blue), ecRNH (light blue), ctRNH (magenta), ttRNH (red), hsRNH (purple). Right: ecRNH (light blue), XMRV WT (green), XMRV ∆C (yellow), HIV (brown). All simulations were carried out at 300K in the AMBER99SB force field with TIP3P water with structures protonated to reflect a pH of 5.5.\n\nIn order to better understand the relationships between dynamic processes in RNase H domains of retroviral origin compared to those from cellular organisms, additional simulations in the absence of divalent ions were performed on a set of retroviral RNase H homologs. In brief, no significant differences are observed between simulations initiated from the XMRV full-length structure compared to its ∆C mutant (in which helix C and the handle region are removed), between the XMRV ∆C mutant compared to the HIV homolog (which naturally lacks the handle sequence), or between any of the retroviral domains compared to ecRNH (Figure 8). This result suggests that the preorganization of the active site on the ps-ns timescale is not significantly altered by differences in amino acid sequence, but rather is inherently imposed by the overall protein fold.\n\n\nConclusions\n\nIn this work we aimed to use molecular dynamics simulations to understand the dynamic behavior of the RNase H family in complex with catalytically required Mg2+ ions. We observe that the well-studied RNase H homolog from E. coli contains a conformationally preorganized active site that is highly rigid on the ps-ns timescale in the presence of a single Mg2+ ion, which is likely located at the B site crystallographically identified by examining the Mn2+ complex. Additionally, we examined the apo state dynamics of the active site—previously validated by comparison to NMR data in the case of the E. coli homolog24—and found that similar patterns of active-site rigidity are present in all homologs examined, including representatives of bacterial, eukaryotic, and retroviral RNases H. This result suggests that active site dynamics are only subtly modulated by amino acid sequence and are primarily imposed by the characteristic protein fold. Although it has long been recognized that RNases H share similar topologies and active-site conformations with other endonucleases46, the present work extends this observation from static crystal structures to dynamics on the ps-ns timescale.\n\nSimulations of the Homo sapiens RNase H homolog and the Thermus thermophilus argonaute protein (a distant RNase H homolog with similar active-site architecture) in complex with two Mg2+ ions and substrate analogs also observe high active-site rigidity47; this is consistent with the present data and implies that active-site preorganization is a general property of this larger family of nucleases. Although selective inhibitors of the HIV RNase H domain have been developed based on the hypothesis that the metal ion dependence of the HIV domain’s catalytic mechanism differs from that of the human homolog15, it is likely that the physical origin of this selectivity is not dependent on the active-site conformations sampled on the ps-ns timescale.\n\n\nData availability\n\nZENODO: Molecular dynamics derived side chain order parameters for Asp, Glu, Asn, and Gln residues in ribonucleases H, and molecular dynamics trajectories for E. coli ribonuclease H. doi:10.5281/zenodo.843148", "appendix": "Author contributions\n\n\n\nKAS: Conceived and designed experiments, performed experiments, analyzed data, wrote the paper. AGP: Conceived and designed experiments, wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nThis work was funded by an NSF graduate research fellowship (KAS) and NIH grant GM50291 (AGP).\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 the Center for Computational Biology and Bioinformatics (C2B2) for computational resources. We thank Jae-hyun Cho and Paul Robustelli for helpful discussions.\n\n\nReferences\n\nTadokoro T, Kanaya S: Ribonuclease H: molecular diversities, substrate binding domains, and catalytic mechanism of the prokaryotic enzymes. FEBS J. 2009; 276(6): 1482–1493. PubMed Abstract \| Publisher Full Text\n\nKanaya S, Kohara A, Miura Y, et al.: Identification of the amino acid residues involved in an active site of Escherichia coli ribonuclease H by site-directed mutagenesis. J Biol Chem. 1990; 265(8): 4615–4621. PubMed Abstract\n\nOda Y, Yamazaki T, Nagayama K, et al.: Individual ionization constants of all the carboxyl groups in ribonuclease HI from Escherichia coli determined by NMR. Biochemistry. 1994; 33(17): 5275–5284. PubMed Abstract \| Publisher Full Text\n\nKanaya S, Itaya M: Expression, purification, and characterization of a recombinant ribonuclease H from Thermus thermophilus HB8. J Biol Chem. 1992; 267(14): 10184–10192. PubMed Abstract\n\nHuang HW, Cowan JA: Metallobiochemistry of the magnesium ion. Characterization of the essential metal-binding site in Escherichia coli ribonuclease H. Eur J Biochem. 1994; 219(1–2): 253–260. PubMed Abstract \| Publisher Full Text\n\nKatayanagi K, Okumura M, Morikawa K: Crystal structure of Escherichia coli RNase HI in complex with Mg2+ at 2.8 A resolution: proof for a single Mg(2+)-binding site. Proteins. 1993; 17(4): 337–346. PubMed Abstract \| Publisher Full Text\n\nGoedken ER, Marqusee S: Co-crystal of Escherichia coli RNase HI with Mn2+ ions reveals two divalent metals bound in the active site. J Biol Chem. 2001; 276(10): 7266–7271. PubMed Abstract \| Publisher Full Text\n\nTsunaka Y, Takano K, Matsumura H, et al.: Identification of single Mn(2+) binding sites required for activation of the mutant proteins of E. coli RNase HI at Glu48 and/or Asp134 by X-ray crystallography. J Mol Biol. 2005; 345(5): 1171–1183. PubMed Abstract \| Publisher Full Text\n\nTsunaka Y, Haruki M, Morikawa M, et al.: Dispensability of glutamic acid 48 and aspartic acid 134 for Mn2+-dependent activity of Escherichia coli ribonuclease HI. Biochemistry. 2003; 42(11): 3366–3374. PubMed Abstract \| Publisher Full Text\n\nNowotny M, Gaidamakov SA, Crouch RJ, et al.: Crystal structures of RNase H bound to an RNA/DNA hybrid: Substrate specificity and metal-dependent catalysis. Cell. 2005; 121(7): 1005–1016. PubMed Abstract \| Publisher Full Text\n\nNowotny M, Gaidamakov SA, Ghirlando R, et al.: Structure of human RNase H1 complexed with an RNA/DNA hybrid: insight into HIV reverse transcription. Mol Cell. 2007; 28(2): 264–276. PubMed Abstract \| Publisher Full Text\n\nOda Y, Nakamura H, Kanaya S, et al.: Binding of metal ions to E. coli RNase HI observed by 1H–15N heteronuclear 2D NMR. J Biomol NMR. 1991; 1(3): 247–255. PubMed Abstract \| Publisher Full Text\n\nKeck JL, Goedken ER, Marqusee S: Activation/attenuation model for RNase H: A one-metal mechanism with second-metal inhibition. J Biol Chem. 1998; 273(51): 34128–34133. PubMed Abstract \| Publisher Full Text\n\nOda Y, Yoshida M, Kanaya S: Role of histidine 124 in the catalytic function of ribonuclease HI from Escherichia coli. J Biol Chem. 1993; 268(1): 88–92. PubMed Abstract\n\nKlumpp K, Hang JQ, Rajendran S, et al.: Two-metal ion mechanism of RNA cleavage by HIV RNase H and mechanism-based design of selective HIV RNase H inhibitors. Nucleic Acids Res. 2003; 31(23): 6852–6859. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNowotny M, Yang W: Stepwise analyses of metal ions in RNase H catalysis from substrate destabilization to product release. EMBO J. 2006; 25(9): 1924–1933. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Vivo M, Dal Peraro M, Klein ML: Phosphodiester cleavage in ribonuclease H occurs via an associative two-metal-aided catalytic mechanism. J Am Chem Soc. 2008; 130(33): 10955–10962. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRosta E, Woodcock HL, Brooks BR, et al.: Artificial reaction coordinate “tunneling” in free-energy calculations: the catalytic reaction of RNase H. J Comput Chem. 2009; 30(11): 1634–1641. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRosta E, Nowotny M, Yang W, et al.: Catalytic mechanism of RNA backbone cleavage by ribonuclease H from quantum mechanics/molecular mechanics simulations. J Am Chem Soc. 2011; 133(23): 8934–8941. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nElsässer B, Fels G: Atomistic details of the associative phosphodiester cleavage in human ribonuclease H. Phys Chem Chem Phys. 2010; 12(36): 11081–8. PubMed Abstract \| Publisher Full Text\n\nHostomsky Z, Hostomska Z, Matthews DA: Ribonucleases H. In Stuart M. Linn and Richard J. Roberts, editors, Nucleases, pages 341–376. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 2nd edition, 1993. Publisher Full Text\n\nIlina T, LaBarge K, Sarafianos SG, et al.: Inhibitors of HIV-1 Reverse Transcriptase-Associated Ribonuclease H Activity. Biology. 2012; 1(3): 521–541. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavies JF 2nd, Hostomska Z, Hostomsky Z, et al.: Crystal structure of the ribonuclease H domain of HIV-1 reverse transcriptase. Science. 1991; 252(5002): 88–95. PubMed Abstract \| Publisher Full Text\n\nStafford KA, Ferrage F, Cho JH, et al.: Side chain dynamics of carboxyl and carbonyl groups in the catalytic function of Escherichia coli ribonuclease H. J Am Chem Soc. 2013; 135(48): 18024–18027. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOda Y, Iwa S, Ohtsuka E, et al.: Binding of nucleic acids to E. coli RNase HI observed by NMR and CD spectroscopy. Nucleic Acids Res. 1993; 21(20): 4690–4695. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnandakrishnan R, Aguilar B, Onufriev AV: H++ 3.0: automating pK prediction and the preparation of biomolecular structures for atomistic molecular modeling and simulations. Nucleic Acids Res. 2012; 40(Web Server issue): W537–W541. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButterwick JA, Patrick Loria J, Astrof NS, et al.: Multiple time scale backbone dynamics of homologous thermophilic and mesophilic ribonuclease HI enzymes. J Mol Biol. 2004; 339(4): 855–871. PubMed Abstract \| Publisher Full Text\n\nButterwick JA, Palmer AG: An inserted Gly residue fine tunes dynamics between mesophilic and thermophilic ribonucleases H. Protein Sci. 2006; 15(12): 2697–2707. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStafford KA, Robustelli P, Palmer AG 3rd: Thermal adaptation of conformational dynamics in ribonuclease H. PLoS Comput Biol. 2013; 9(10): e1003218. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBowers KJ, Chow E, Xu H, et al.: Scalable algorithms for molecular dynamics simulations on commodity clusters. In Proceedings of the 2006 ACM/IEEE conference on Supercomputing, page 84, Tampa, Florida, 2006. ACM. Publisher Full Text\n\nHornak V, Abel R, Okur A, et al.: Comparison of multiple Amber force fields and development of improved protein backbone parameters. Proteins. 2006; 65(3): 712–725. PubMed Abstract \| Publisher Full Text\n\nJorgensen WL, Chandrasekhar J, Madura JD, et al.: Comparison of simple potential functions for simulating liquid water. J Chem Phys. 1983; 79(2): 926–935. Publisher Full Text\n\nKräutler V, van Gunsteren WF, Hünenberger PH: A fast SHAKE algorithm to solve distance constraint equations for small molecules in molecular dynamics simulations. J Comput Chem. 2001; 22(5): 501–508. Publisher Full Text\n\nAaqvist J: Ion-water interaction potentials derived from free energy perturbation simulations. J Phys Chem. 1990; 94(21): 8021–8024. Publisher Full Text\n\nRobustelli P, Stafford KA, Palmer AG 3rd: Interpreting protein structural dynamics from NMR chemical shifts. J Am Chem Soc. 2012; 134(14): 6365–6374. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChandrasekhar I, Clore GM, Szabo A, et al.: A 500 ps molecular dynamics simulation study of interleukin-1 beta in water. Correlation with nuclear magnetic resonance spectroscopy and crystallography. J Mol Biol. 1992; 226(1): 239–250. PubMed Abstract \| Publisher Full Text\n\nKatayanagi K, Miyagawa M, Matsushima M, et al.: Structural details of ribonuclease H from Escherichia coli as refined to an atomic resolution. J Mol Biol. 1992; 223(4): 1029–1052. PubMed Abstract \| Publisher Full Text\n\nIshikawa K, Okumura M, Katayanagi K, et al.: Crystal structure of ribonuclease H from Thermus thermophilus HB8 refined at 2.8 A resolution. J Mol Biol. 1993; 230(2): 529–542. PubMed Abstract \| Publisher Full Text\n\nTadokoro T, You DJ, Abe Y, et al.: Structural, thermodynamic, and mutational analyses of a psychrotrophic RNase HI. Biochemistry. 2007; 46(25): 7460–7468. PubMed Abstract \| Publisher Full Text\n\nRatcliff K, Corn J, Marqusee S: Structure, stability, and folding of ribonuclease H1 from the moderately thermophilic Chlorobium tepidum: comparison with thermophilic and mesophilic homologues. Biochemistry. 2009; 48(25): 5890–5898. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHimmel DM, Maegley KA, Pauly TA, et al.: Structure of HIV-1 reverse transcriptase with the inhibitor beta-Thujaplicinol bound at the RNase H active site. Structure. 2009; 17(12): 1625–1635. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhou D, Chung S, Miller M, et al.: Crystal structures of the reverse transcriptase-associated ribonuclease H domain of xenotropic murine leukemia-virus related virus. J Struct Biol. 2012; 177(3): 638–645. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKirby KA, Marchand B, Ong YT, et al.: Structural and inhibition studies of the RNase H function of xenotropic murine leukemia virus-related virus reverse transcriptase. Antimicrob Agents Chemother. 2012; 56(4): 2048–2061. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMamatkulov S, Fyta M, Netz RR: Force fields for divalent cations based on single-ion and ion-pair properties. J Chem Phys. 2013; 138(2): 024505. PubMed Abstract \| Publisher Full Text\n\nHaruki M, Noguchi E, Nakai C, et al.: Investigating the role of conserved residue Asp134 in Escherichia coli ribonuclease HI by site-directed random mutagenesis. Eur J Biochem. 1994; 220(2): 623–631. PubMed Abstract \| Publisher Full Text\n\nYang W, Steitz TA: Recombining the structures of HIV integrase, RuvC and RNase H. Structure. 1995; 3(2): 131–134. PubMed Abstract \| Publisher Full Text\n\nMaláč K, Barvík I: Complex between human RNase HI and the phosphonate-DNA/RNA duplex: molecular dynamics study. J Mol Graph Model. 2013; 44: 81–90. PubMed Abstract \| Publisher Full Text\n\nStafford KA, Palmer AG III: Molecular dynamics analysis of side chain order parameters for asp glu, asn and gln residues in ribonucleases H. ZENODO. 2014. Data Source" }	[ { "id": "4005", "date": "21 Mar 2014", "name": "Rafael Najmanovich", "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 set out to use molecular dynamics simulations to the binding of a required magnesium ion on Ribonuclease H1, utilizing a number of distinct structures including those crystallized in different conditions or forms (Apo or different types of Holo forms) and orthologs. The main result found by the authors is that the binding site is very rigid and pre-organized in the presence of a single magnesium ion. The authors show that this observation holds for all orthologs studied. Based on this, the authors suggest that 'the active site dynamics are only subtly modulated by amino acid sequence and are primarily imposed by the characteristic protein fold'. It is indeed interesting that most RNases not only conserve the structure of the binding site but also dynamic properties in the ns-ps range. However, I feel that showing a multiple sequence alignment of the RNase H family highlighting the members used in this study as well as the positions that constitute the substrate and metal binding sites would considerably help understand that indeed this pre-organization comes about despite divergent binding-site sequences. Without seeing such an alignment is hard to accept that the rigidity of the binding-site is not affected by changes in binding-site sequence. If the binding-site sequences turn out to be highly conserved upon analysis of the multiple sequence alignment, it may be necessary to perform mutations in silico in the binding-site to support the conclusion above.", "responses": [ { "c_id": "874", "date": "20 Jun 2014", "name": "Kate Stafford", "role": "Author Response", "response": "Thank you for your thoughtful comments on our paper. In response to your suggestion, we have published a version 2 of the paper containing a new Figure 8, which shows a structure-based multiple sequence alignment for all RNases H examined (see reviewer response above). The positions of the active-site residues and the substrate-binding regions have been highlighted on the MSA to facilitate comparison of the sequences within these regions. The point you raise can be subtle - since some sequence similarity is required for maintenance of the overall fold, rather than for specific substrate interactions - so we have edited the text of the Conclusions section for clarity and changed the final sentence of the Abstract to read \"...may primarily be imposed...\". We have also updated the Conclusions section with a new reference to a recent paper on the evolution of the RNase H superfamily (ref 50)." } ] }, { "id": "4492", "date": "20 May 2014", "name": "Gaetano Montelione", "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 exciting paper which beautifully demonstrates the value of MD simulations coupled with experimental data for understanding structure-dynamic-function mechanisms. The research appears to be carefully executed, with proper controls, and the paper is well written. Together with related papers (e.g, Stafford et al., JACS 2014 - ref 24) it would be an excellent topic paper in the area of protein structure and function for an undergraduate or graduate seminar course. The results indicate that the apo state of Ec RNase H is conformationally preorganized to favor binding of magnesium ion. MD simulations suggest the primary Mg+2-binding site is actually the observed Mn+2 –binding site B. This is an interesting application of MD to assess the significance of various metal-binding sites previously observed in X-ray crystal structures. 13C nuclear relaxation measurements reveal no statistically significant differences in the fast-time-scale side-chain dynamics of DENQ residues for complexes with Mg+2 modelled into metal binding sites compared with apo Ec RNase H. However, these modelling results suggest some minor structural changes in the active site that may enhance subsequent substrate binding. MD simulations from a phylogenetically diverse set of RNase H proteins, including representatives of bacterial, eukaryotic, and retroviral RNAses H, suggest that active sites that are conformationally-preorganized for metal binding are a common feature of the RNase H domain family. The work is suitable for publication in F1000Research but I do suggest minor revisions to address the following points:Fig. 1 shows strong conservation of active site residues among bacterial homogues, and Fig. 8 documents similarity in active site dynamics across a broad phylogenetic distribution, as characterized by MD simulations. The authors conclude that these dynamic properties are only subtly modulated by amino acid sequence and primarily imposed by the characteristic protein fold. As suggested by Reviewer 1 (Rafael Najmanovich), it would be helpful to document the active-site sequence variation across these homologs to support this conclusion. The authors state “all simulations were performed at a pH of 5.5 to recapitulate the conditions used in prior NMR experiments on ecRNH”. What is the pH optimum of enzyme function? Is there any information about “physiological pH”. Based on the pKa’s of active site ionizable groups, how might the pH selected for this study impact the conclusions.", "responses": [ { "c_id": "873", "date": "20 Jun 2014", "name": "Kate Stafford", "role": "Author Response", "response": "Thank you for your thoughtful comments on our paper. In response to your suggestions, we have published a version 2 of the paper which addresses both of your points, specifically:A new Figure 8 has been added to the paper, showing a structure-based sequence alignment for all of the bacterial (soRNH, ecRNH, ctRNH, ttRNH), eukaryotic (hsRNH), and retroviral (XMRV, HIV) RNase H structures used to initiate simulations as part of this work. The pH optimum of the RNase H reaction was previously mentioned only in the Introduction (refs 3 and 4). We have added a comment in the Methods section on the motivation for the choice of pH 5.5 in the NMR experiments and the impact of this on the simulations. In brief, the active site residues are expected to be unprotonated in the presence of Mg2+ at both the experimental pH of 5.5 and the reaction optima of 7.5-9.5." } ] } ]	1	https://f1000research.com/articles/3-67
https://f1000research.com/articles/3-79/v1	27 Mar 14	{ "type": "Case Report", "title": "Severe form of hemolytic-uremic syndrome with multiple organ failure in a child: a case report", "authors": [ "Dino Mijatovic", "Ana Blagaic", "Zeljko Zupan", "Ana Blagaic", "Zeljko Zupan" ], "abstract": "Introduction: Hemolytic-uremic syndrome (HUS) is a leading cause of acute renal failure in infants and young children. It is traditionally defined as a triad of acute renal failure, hemolytic anemia and thrombocytopenia that occur within a week after prodromal hemorrhagic enterocolitis. Severe cases can also be presented by acute respiratory distress syndrome (ARDS), toxic megacolon with ileus, pancreatitis, central nervous system (CNS) disorders and multiple organ failure (MOF).Case presentation: A previously healthy 4-year old Caucasian girl developed acute renal failure, thrombocytopenia and hemolytic anemia following a short episode of abdominal pain and bloody diarrhea. In the next week of, what initially appeared as typical HUS, she developed MOF, including ileus, pancreatitis, hepatitis, coma and ARDS, accompanied by hemodynamic instability and extreme leukocytosis. Nonetheless, the girl made a complete recovery after one month of the disease. She was successfully treated in the intensive care unit and significant improvement was noticed after plasmapheresis and continuous veno-venous hemodialysis.Conclusions: Early start of plasmapheresis and meticulous supportive treatment in the intensive care unit, including renal placement therapy, may be the therapy of choice in severe cases of HUS presented by MOF. Monitoring of prognostic factors is important for early performance of appropriate diagnostic and therapeutical interventions.", "keywords": [ "Hemolytic-uremic syndrome (HUS) can be classified as typical", "usually when provoked by Shiga-toxin (STx) produced by enterohemorrhagic Escherichia coli serotype H157:O7", "or atypical", "when triggered by other microbes’ antigens and toxins1. The overall incidence of STx-HUS is estimated to be 2.1 per 100 000 people", "with a peak incidence in children younger than 5 years (6.1 per 100 000)1. Approximately two-thirds of patients have to be dialyzed2. Clinical manifestations of HUS are the consequence of severe systemic inflammation and immune reactions that lead to thrombotic microangiopathy in susceptible organs. In typical HUS", "those reactions are initiated by binding of STx to the endothelial membrane-bound Gb3 receptor and subsequent activation of platelets and leukocytes3", "while in atypical HUS", "those reactions are mediated by excessive activation of complement. Clinical manifestations can overlap with another similar syndrome – thrombotic thrombocytopenic purpura (TTP)5. Typical HUS has better prognosis in terms of morbidity and mortality than atypical and severe forms of HUS complicated by multiple organ failure (MOF)1", "4." ], "content": "Introduction\n\nHemolytic-uremic syndrome (HUS) can be classified as typical, usually when provoked by Shiga-toxin (STx) produced by enterohemorrhagic Escherichia coli serotype H157:O7, or atypical, when triggered by other microbes’ antigens and toxins1. The overall incidence of STx-HUS is estimated to be 2.1 per 100 000 people, with a peak incidence in children younger than 5 years (6.1 per 100 000)1. Approximately two-thirds of patients have to be dialyzed2. Clinical manifestations of HUS are the consequence of severe systemic inflammation and immune reactions that lead to thrombotic microangiopathy in susceptible organs. In typical HUS, those reactions are initiated by binding of STx to the endothelial membrane-bound Gb3 receptor and subsequent activation of platelets and leukocytes3, while in atypical HUS, those reactions are mediated by excessive activation of complement. Clinical manifestations can overlap with another similar syndrome – thrombotic thrombocytopenic purpura (TTP)5. Typical HUS has better prognosis in terms of morbidity and mortality than atypical and severe forms of HUS complicated by multiple organ failure (MOF)1,4.\n\n\nCase description\n\nIn 2012, a 4-year old Caucasian girl, on holiday with her parents, was brought to University Hospital Rijeka emergency department with abdominal cramps, vomiting and bloody diarrhea that had started 24 hours before her admission. Gastroenterocolitis was initially suspected, but as the abdominal pain intensified, an explorative laparotomy was performed on the third day of disease to exclude perforated appendicitis, which the surgeon eventually did not confirm. On the first postoperative day she appeared confused, dyspnoic, hypotensive, pale, oliguric and her laboratory findings were: hemoglobin (Hb) 60 g/L, hematocrit (Hct) 19%, platelet count 43×109/L, white blood cell (WBC) count 17×109/L, urea 16.6 mmol/L, creatinine 308 μmol/L, creatine kinase 971 U/L, aspartate aminotransferase (AST) 317 U/L, alanine transaminase (ALT) 276 U/L, C-reactive protein (CRP) 175.2 mg/L, lactate dehydrogenase (LDH) 3856 U/L, albumins 19 g/L, complement component 3 (C3) 0.5 g/L, complement component 4 (C4) 0.08 g/L, degree of in vivo hemolysis 50 mg/L, pH 7.265, pCO2 3.64 kPa, pO2 9.2 kPa, HCO3 12 mmol/l, SBE -13.1 mmol/L. She was immediately admitted to intensive care where she was analgosedated with 6 mg of midazolam and 200 mcg of fentanyl, intubated with an uncuffed orotracheal tubus and invasive mechanical ventilation was started. During the first week of disease she received multiple transfusions of packed red blood cells, fresh frozen plasma and platelets as continuous adrenergic and diuretic support was established with 5–8 mcg/kg/min of dopamine and 50 mg/24 h of furosemide, respectively. After additional laboratory tests, typical HUS caused by reaction to E. coli toxin was diagnosed. By the end of the first week of disease her medical condition had not improved and she became anuric, accompanied by combined metabolic and respiratory acidosis, pulmonary edema, bilateral pleural and pericardial effusions, ascites and persistent anemia and thrombocytopenia. Continuous veno-venous hemodiafiltration (CVVHDF) as well as plasmapheresis were indicated. Access was established via an intravascular catheter placed through the right internal jugular vein guided by ultrasound. CVVHDF was performed for 12 days. Characteristics of dialysis included a dialyzer membrane surface area of 0.8 m2, solution containing bicarbonate (multiBic 2 mmol/L potassium solution for haemofiltration; Fresenius Medical Care Deutschland GmbH) and a blood flow rate of 3–6 ml/kg/min. The system was anticoagulated with 100 IU/kg of unfractionated heparin each day. An additional blood flow rate for ultrafiltration was 0.5–2 ml/kg/h. The renal replacement therapy was only stopped for 3 hours a day in order to carry out plasmapheresis which was also initiated on the 8th day of disease. This lasted for 10 days, for 2 hours each day and 40 ml/kg of fresh frozen plasma was used per day. During this period continuous analgosedation with 0.15 mg/kg/h midazolam and 0.075 mg/kg/h morphine, as well as invasive mechanical ventilation were maintained. Continuous adrenergic support was stopped after the 6th day of disease. After 12 days of renal placement therapy, diuresis started to improve while serum creatinine and urea started to decrease. Despite the polyuria during the initial phase of recovery of renal function, because of the still significant fluid retention, we established negative fluid balance by continuous infusion of furosemid. Renal function had completely recovered by the 25th day of disease (Figure 1). We have reason to believe that the blood levels of creatinine and urea at the time that the CVVHDF was started were falsely lower because of hemodilution caused by the body’s extreme retention of extracellular water.\n\nPlasma creatinine and urea levels and urine output during the disease course and applied treatment show significant improvement of renal function by the end of CVVHD and plasmapheresis and normal function is restored in the 4th week of disease without the evidence of permanent damage.\n\nDuring the course of plasmapheresis a marked recovery of the platelet count and hemoglobin level was noticed, as well as a significant decline of LDH levels and the degree of in vivo hemolysis. After termination of plasmapheresis, an additional time was needed for restoration of platelets and normal erythrocytes to optimal levels because of their biological cell cycle. We were restrictive throughout in the transfusion of packed red cells, aiming for a target hemoglobin level of 70–80 g/L.\n\nThe most striking laboratory finding during the first two weeks of disease was extreme leukocytosis which persisted during the second week of disease; the highest WBC count was 94×109/L. As no other signs of systemic bacterial infection or invasive mycosis were noticed, we believe that leukocytosis was provoked by the systemic inflammatory response to toxemia. We observed the gradual decrease in the WBC count after the initiation of plasmapheresis and its subsequent normalization by the end of the third week of disease. This trend was correlated with the values of LDH plasma levels (Figure 2).\n\nPlasmapheresis enabled the clearance of proinflammatory factors as the WBC and LDH level normalized during and after this course of treatment. Red lines denote upper reference limit in our laboratory - for plasma LDH 485 IU/L and WBC 10×109/L.\n\nBilateral pleural effusions needed evacuation by thoracocentesis on two occasions, on the 8th and 13th day of disease. A CT scan of the thorax performed on the 15th day of disease revealed a massive left-sided hemothorax complicated by compressive and obstructive atelectasis of the left lung (Figure 3). The hemothorax had to be surgically drained on the 16th day of disease. During the second week of disease we confirmed acute respiratory distress syndrome (ARDS) according to PaO2 to FiO2 ratio and evidence of hepatization of the lungs on serial ultrasound scans. Because of the pulmonary manifestations of the disease, a protective ventilation strategy using was employed. We mostly used airway pressure release ventilation (APRV) and biphasic positive airway pressure (BIPAP) modes on a Dräger Evita XL mechanical ventilator and frequently applied lung recruitment with the positive end-expiratory pressure (PEEP) technique monitored by ultrasound. Respiratory function completely recovered by the end of the third week of disease without apparent permanent damage and the girl was successfully extubated on the 24th day of disease.\n\nWe ceased the analgosedation on the 12th day of disease in order to evaluate the neurological status of the girl. She appeared comatose, estimated as GCS 4 (E2V1M1) on the Glasgow Coma Scale with flaccid tetraparesis, rotatory nystagmus and symmetrical mydriatic pupils with slow response to light stimulation. An MRI scan of the brain recorded on the 12th day of disease revealed areas of raised signal in the pons, bilaterally in lateral parts of the thalamus (Figure 4), as well as in the right external capsule and the left internal capsule. These findings were analogous with lesions already described in literature believed to be the result of microvascular damage8. Analgosedation was reinstated and afterwards periodically discontinued every few days for approximately 6 hours in order to examine the girl’s clinical neurological status. After the 18th day her level of consciousness and neurological status started to gradually improve and we decided to definitively cease analgosedation 7 days later. No permanent neurological damage was recorded.\n\nWe were concerned by severe gastric retention, absent or silent peristalsis and constipation with episodical soft and rare stools during the second week of disease. An MRI of the abdomen performed on the 12th day of disease showed an edematous intestinal wall and distension of the jejunum, ileum and colon ascendens with several air fluid levels (Figure 5). Laboratory findings displayed elevated serum transaminases, total and direct bilirubin and amylase as well as hypoalbuminemia. After prodromal gastroenterocolitis, paralytic ileus obviously evolved and was persistent throughout the second and third week of disease. Therefore, all necessary fluid, macronutrients and micronutrients were completely delivered parenterally via a central venous catheter from the start of disease. In combination with already initiated parenteral nutrition, we carefully started enteral feeding by the end of the second week of disease. As no signs of intraabdominal compartment syndrome, perforation or peritonitis were registered, we wanted to reestablish gastrointestinal function conservatively. Therefore, we applied continuous infusion of 6–10 mg/24h of metoclopramid and intermittent rectal suppositories of 2.5–5 mg of bisacodyl for the next two weeks. We also progressively increased the contribution of the enteral input as soon as peristalsis started to improve. Bowel function slowly recovered by the beginning of the fourth week of disease.\n\nAs the patient and her family had been on holiday at the time of admission, an in-person follow up was not possible. However, the authors made contact with the family in 2013 and the parents reported that their daughter was perfectly healthy and without apparent consequences.\n\n\nDiscussion and conclusion\n\nHere we presented the case of a 4-year old girl who developed a severe form of typical HUS which initially presented classically, but rapidly progressed into MOF which severely affected the central nervous system, renal, gastrointestinal and cardiovascular function, and induced ARDS and hematological disorders.\n\nDiarrhoea-associated HUS is the most severe clinical manifestation of infection with Shiga toxin-producing Escherichia coli and is more common in children1. Pathogenesis of the syndrome is based on the reaction of the innate immune system to toxemia3. Characteristic features of the syndrome are hemorrhagic enterocolitis, hemolytic anemia, thrombocytopenia and acute renal failure, but some patients develop more unusual manifestations that potentially lead to MOF and increase mortality4. Even though 70% of children with HUS recover without permanent health consequences, 2–5% of patients die in the acute phase1,2.\n\nAmong less usual manifestations, CNS involvement is the most frequent and one that significantly increases morbidity and mortality6. Neurological signs range from epileptic seizures to reduced consciousness level and focal motor deficits1,7, but are mostly the result of temporary dysfunction rather than irreversible damage. Brain MRI is the diagnostic method of choice when analyzing possible CNS lesions and usually reveals a pattern of symmetrical hyperintensities in basal ganglia and the thalamus8.\n\nThe gastrointestinal and hepatobiliary tract may be affected from the esophagus to the perianal area and possible disorders include ileus, intussusception, bowel distension, perforation, necrosis, toxic megacolon, intestinal stricture, rectal prolapse, hepatocellular cholestasis and pancreatitis, which may lead to diabetes mellitus4,9.\n\nCardiovascular instability can be a consequence of tachyarrhythmia caused by HUS, electrolyte imbalance and toxic myocarditis with subsequent dilatative cardiomyopathy4,6.\n\nThe mainstay of treatment is supportive and includes control of fluid and electrolyte balance, optimal enteral and parenteral nutrition, use of hemodialysis (required in approximately two-thirds of patients2), control of hemodynamic stability and judicious transfusion of blood derivatives. Plasmapheresis with fresh frozen plasma is reserved for the most severe cases10, although it was proven ineffective in some recent controlled clinical trials11. Platelet transfusions are avoided and limited for control of active bleeding, considering some studies have suggested it could contribute to the microthrombosis and worsen outcome4. Use of antibiotics is controversial and should be avoided, since some studies have shown them to be harmful as possible triggers for the development of HUS in patients with enterohemorrhagic E. coli infection1.\n\nAlready known predictors of poor outcome, including death and chronic renal and lung disease, are prolonged oliguria or anuria, need for hemodialysis, neurological impairment, persistent leukocytosis > 20×109/L, hematocrit < 23% on admission and severe dysfunction of the gastrointestinal system6,11. The recorded extreme leukocytosis in this case was related to the particularly high activity of systemic inflammation as a result of toxemia rather than infection. It could be an indicator of HUS complicated with MOF and, therefore, one of the main negative prognostic factors2,3,6. Despite the fact that all these predictors were recognized in the acute stage of disease, our patient fully recovered without any apparent sequelae during the follow-up period of one year. This case indicates that early diagnosis, thorough supportive treatment, including renal replacement therapy and early plasmapheresis, are crucial interventions for favorable outcomes in severe cases of typical HUS presented by MOF.\n\n\nConsent\n\nWritten informed consent for publication of this case report and corresponding images was obtained from the patient’s parents.", "appendix": "Author contributions\n\n\n\nAll three authors analyzed the data and critically revised the manuscript. DM is the first author who designed and wrote the majority of the paper and collected the data. AB made images and graphical analysis of the data. ŽŽ was the primary clinician responsible for the treatment of this patient and had the initial idea about the paper.\n\n\nCompeting interests\n\n\n\nNo competing interest were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nNoris M, Remuzzi G: Hemolytic uremic syndrome. J Am Soc Nephrol. 2005; 16(4): 1035–50. PubMed Abstract \| Publisher Full Text\n\nGerber A, Karch H, Allerberger F, et al.: Clinical course and the role of Shiga toxin-producing Escherichia coli infection in the hemolytic-uremic syndrome in pediatric patients, 1997–2000, in Germany and Austria: a prospective study. J Infect Dis. 2002; 186(4): 493–500. PubMed Abstract \| Publisher Full Text\n\nExeni RA, Fernández GC, Palermo MS: Role of polymorphonuclear leukocytes in the pathophysiology of typical hemolytic uremic syndrome. ScientificWorldJournal. 2007; 7: 1155–64. PubMed Abstract \| Publisher Full Text\n\nScheiring J, Andreoli SP, Zimmerhackl LB: Treatment and outcome of Shiga-toxin-associated hemolytic uremic syndrome (HUS). Pediatr Nephrol. 2008; 23(10): 1749–60. PubMed Abstract \| Publisher Full Text\n\nGeorge JN, Al-Nouri ZL: Diagnostic and therapeutic challenges in the thrombotic thrombocytopenic purpura and hemolytic uremic syndromes. Hematology Am Soc Hematol Educ Program. 2012; 2012(1): 604–609. PubMed Abstract \| Publisher Full Text\n\nOakes RS, Siegler RL, McReynolds MA, et al.: Predictors of fatality in postdiarrheal hemolytic uremic syndrome. Pediatrics. 2006; 117(5): 1656–62. PubMed Abstract \| Publisher Full Text\n\nNathanson S, Kwon T, Elmaleh M, et al.: Acute neurological involvement in diarrhea-associated hemolytic uremic syndrome. Clin J Am Soc Nephrol. 2010; 5(7): 1218–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSteinborn M, Leiz S, Rüdisser K, et al.: CT and MRI in haemolytic uraemic syndrome with central nervous system involvement: distribution of lesions and prognostic value of imaging findings. Pediatr Radiol. 2004; 34(10): 805–10. PubMed Abstract \| Publisher Full Text\n\nRiley MR, Lee KK: Escherichia coli O157:H7–associated hemolytic uremic syndrome and acute hepatocellular cholestasis: a case report. J Pediatr Gastroenterol Nutr. 2004; 38(3): 352–354. PubMed Abstract\n\nBambauer R, Latza R, Schiel R: Therapeutic apheresis in the treatment of hemolytic uremic syndrome in view of pathophysiological aspects. Ther Apher Dial. 2011; 15(1): 10–19. PubMed Abstract \| Publisher Full Text\n\nGarg AX, Suri RS, Barrowman N, et al.: Long-term renal prognosis of diarrhea-associated hemolytic uremic syndrome: A systematic review, meta-analysis, and meta-regression. JAMA. 2003; 290(10): 1360–70. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4458", "date": "17 Apr 2014", "name": "Zeljko Bosnjak", "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 a very interesting but extremely difficult case report that many will find quite helpful. I have few minor suggestions: Abstract (case presentation): rewrite the second sentence.Figures 1 and 2: Could easily be combined into one figure, with single labeling on the top, and placing descriptions of the parameters on top of each tracing. Also the units of diuresis should be changed to per day instead. For plasma creatinine, include the micro symbol. Figures 3-5: Perhaps white arrows would be a better choice.", "responses": [ { "c_id": "858", "date": "11 Jun 2014", "name": "Dino Mijatovic", "role": "Author Response", "response": "Dear Professor Bošnjak,thank you very much for your comment! Unfortunately, we weren't able to write the response since we have been extremely occupied with organization of the national congress. Anyway, we found ourselves very grateful for your meticulous and thoughtful analysis. We decided to rewrite the second sentence in Case presentation. We also made proper corrections regarding the units of diuresis and plasma creatinine. We have still not combined figures 1 and 2 into one as we want to stress the extreme leukocytosis and its recovery during the PAF treatment, the process that was parallel to the monitored activity of LDH, which is the biomarker of this disease. We believe that this pattern becomes more appreciated when it is stressed in the figure with one or two diagrams. In the uniform picture it would be among 4 different diagrams.We have left black arrows for figures 3-5 because their tops are a greater contrast to the density of the object on the scan." } ] }, { "id": "4457", "date": "30 Apr 2014", "name": "Grant Luxton", "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 case report documenting the response to treatment with best supportive care including dialysis as well as plasmapheresis in a four year old child with \"typical\" haemolytic uraemic syndrome. As the authors point out, trial evidence is not supportive of plasmapheresis for this condition however there may be situations where it may be of benefit - especially in a patient not responding to best supportive care. I have no issues with the case report and believe it is worth indexing to remind us that plasma exchange should still be considered in selected patients with this disease. With the potential availabilty of eculizumab however, the role of plasmapheresis may become even more limited.", "responses": [ { "c_id": "849", "date": "04 Jun 2014", "name": "Dino Mijatovic", "role": "Author Response", "response": "I welcome your comment! We (all the authors) discussed the potential use of eculizumab during the early course of the disease in the second week. However, we didn't have any experience with it and it seemed to be very expensive treatment. Also, we found more data in the recent literature confirming the efficacy of this drug in atypical cases of HUS, rather than typical. Therefore, we decided to initiate the transfusion treatment with full awareness and precautions regarding its potential side-effects, and to leave the eculizumab as the second-line therapy." } ] } ]	1	https://f1000research.com/articles/3-79
https://f1000research.com/articles/3-58/v1	14 Feb 14	{ "type": "Case Report", "title": "Case Report: A case report of acromegaly associated with primary aldosteronism", "authors": [ "Joanna Matrozova", "Silvia Vandeva", "Sabina Zacharieva", "Silvia Vandeva", "Sabina Zacharieva" ], "abstract": "We describe a patient with a rare combination of acromegaly and primary aldosteronism. A 37 year-old female patient was diagnosed with acromegaly on the basis of typical clinical, hormonal and image characteristics. She presented also with one of the most common co-morbidities – arterial hypertension. The patient has been regularly followed-up and after three surgical interventions, irradiation and adjuvant treatment with a dopamine agonist, acromegaly was finally controlled in 2008 (20 years after diagnosis). Arterial hypertension however, remained a therapeutic problem even after prescription of four antihypertensive drugs. She had normal biochemical parameters, except for low potassium levels 3.2 (3.5-5.6) mmol/l. This raised the suspicion of primary hyperaldosteronism, confirmed by a high aldosterone to plasma rennin activity ratio, high aldosterone level after a Captopril challenge test and visualization of a 35 mm left adrenal nodule on a CT scan. After an operation, the patient recovered from hypokalemia and antihypertensive therapy was reduced to a small dose of a Ca blocker.Co-morbid arterial hypertension is common in acromegaly, though it is rare for this to be caused by Conn’s adenoma. The association of Conn’s adenoma with acromegaly has been interpreted in two lines: as a component of multiple endocrine neoplasia type (MEN1) syndrome or as a direct mitogenic effect of hyperactivated GH-IGF1 axis.", "keywords": [ "In the literature several patients with primary aldosteronism (PA) associated with endocrine and non-endocrine tumors have been reported1–3. These cases were attributed mainly to multiple endocrine neoplasia type 1 (MEN1)", "which is an autosomal dominantly inherited condition", "characterized by the association of tumors of the pituitary and the parathyroid glands", "the endocrine pancreas", "the adrenal glands and neuroendocrine tumors4. Aldosterone-producing adenomas have been described in various combinations", "mainly with hyperparathyroidism", "prolactinomas and pancreatic endocrine tumors", "but none of these have been associated with acromegaly2", "3", "5." ], "content": "Introduction\n\nIn the literature several patients with primary aldosteronism (PA) associated with endocrine and non-endocrine tumors have been reported1–3. These cases were attributed mainly to multiple endocrine neoplasia type 1 (MEN1), which is an autosomal dominantly inherited condition, characterized by the association of tumors of the pituitary and the parathyroid glands, the endocrine pancreas, the adrenal glands and neuroendocrine tumors4. Aldosterone-producing adenomas have been described in various combinations, mainly with hyperparathyroidism, prolactinomas and pancreatic endocrine tumors, but none of these have been associated with acromegaly2,3,5.\n\nOn the other hand, the prevalence of adrenal lesions in sporadic cases of acromegaly is higher than in the general population, possibly due to the permissive role of growth hormone (GH) and insulin-like growth factor 1 (IGH1) on tumorigenesis6–8. Previously benign, non-secreting adrenal tumors have been described in patients with acromegaly6–8. However, to our knowledge the association between acromegaly and Conn’s adenoma in sporadic acromegaly has been reported only in isolated cases9,10.\n\nWe describe herein a case of acromegaly, associated with PA due to an aldosterone-producing adenoma of the adrenal gland.\n\n\nCase report\n\nIn May 2008, a 57 year-old, Caucasian woman with a 20 year history of acromegaly was hospitalized in our clinic for a regular follow-up. Acromegaly was diagnosed in 1988 on the basis of a GH of 9.3 mIU/L, which was not suppressed during a glucose tolerance test (OGTT; GH 13.6 mIU/L) and a macroadenoma of the pituitary gland visualized on CT. Prolactin levels were also high: 3300 mIU/l (normal value <650). A non-radical transsphenoidal adenomectomy was performed in the same year and another one in 1991 after growth of the remnant tumor mass. Histology conducted in 1988 showed somatoprolactinoma. In 1993, she had a transcranial adenomectomy due to remnant macroadenoma. As the disease activity (i.e. GH hypersecretion and presence of tumor mass) still persisted, in 1994 she was treated with radiotherapy and in the period between 1994 and 2001 she received dopamine agonists (15 mg daily of bromocriptine for six years and 2×0.5 mg weekly cabergoline for the last four months of the period). Acromegaly was still not controlled and a treatment with a somatostatin analogue (vapreotide s.c. implant, 396 mg every 12 weeks) was started in 2001. The patient was estimated to be a partial responder with no significant improvement in hormonal parameters, so she was switched back to 2×0.5 mg weekly dose of cabergoline. The MRI from 2008 showed a tumor remnant of 18 mm, spreading towards the right part of the sphenoidal sinus.\n\nThe patient had a past medical history of multinodular goiter, operated on in 2001 and recurrent in 2008. She has been hypertensive since she was 37 years old. Her hypertension has never been well controlled on a triple therapy, including a diuretic (enalapril maleate 2×20 mg, nifedipine 4×10 mg and chlortalidone 100 mg daily). She was operated on for a colon polyp in 1994 and was followed-up for myoma of the uterus. Family history of hypertension was recorded with both parents being hypertensive.\n\nOn admission in 2008 she was 1.70 m tall, weighed 83 kg with a waist circumference of 89 cm and had a BMI of 29. Her blood pressure was 150/90 mmHg. Arterial blood pressure monitoring (ABPM) showed a mean daily systolic blood pressure (SBP) of 150 mmHg and diastolic blood pressure (DBP) of 93 mmHg. At that time she was treated with nitrendipine 20 mg twice a day, enalapril 20 mg twice a day, indapamide 2.5 mg per day, prazosin 0.5 mg per day, and cabergoline 2×0.5 mg weekly. Physical examination showed typical acromegalic features such as enlargement of the hands and feet, deep nasolabial folds and macroglossia.\n\nThe association of resistant hypertension and hypokalemia raised the suspicion of PA. Baseline laboratory data are shown in Table 1. Treatment with ACE-inhibitors and diuretics was stopped for 20 days to investigate her Renin-Angiotensin-Aldosterone-System (RAAS) before the next presentation to our clinic. In our unit the aldosterone to renin ratio (ARR) is used as a screening test for PA. In patients with elevated basal ARR (>750 pmol/l per ng/ml/h) a confirmatory test (Captopril challenge test) is performed. The diagnosis of PA is confirmed if the aldosterone is >330 pmol/l at the 90th minute after the oral administration of 50 mg of Captopril, with the patient asked to remain in a sitting position throughout the test. In our patient the basal and post-captopril aldosterone to renin ratio was 1260 and 3545, respectively. Suppressed plasma rennin activity (PRA) was measured at the beginning and the end of the test (<0.2 ng/ml, normal range 0.3–3). Basal aldosterone was 252 pmol/l and 709 pmol/l at the end of the test. The adrenal CT scan showed a 35 mm nodule in the left adrenal gland (Figure 1) and aldosterone-producing adenoma (Conn’s adenoma) was diagnosed. The patient was operated on in August 2008. A tumor of 35 mm was found and histology (hematoxylin and eosin) data showed a tumor of adrenal cortex origin, consisting of light cells.\n\nIGF1- insulin-like growth factor-1, GH- growth hormone, FT4- free thyroxine\n\nThe patient had a follow-up visit three months after the operation. Recovery from hypokalemia was recorded (potassium: 4.95 mmol/l). Over this period the antihypertensive therapy was reduced and she was receiving only a small dose of a Ca blocker (nitrendipine 40 mg daily). Her blood pressure was 130/80 mmHg. ABPM showed a mean daily SBP of 138 mmHg and a DBP of 83 mmHg, and a mean heart rate of 53 bpm. A Captopril test was performed which revealed normal aldosterone levels with a tendency to low renin values: 0 min PRA 0.3 ng/ml/h, aldosterone 82 pmol/l, ARR 273; 90 min PRA 0.3 ng/ml/h, aldosterone 98.9 pmol/l, ARR 329.\n\nIn the last three years of her regular follow-up visits the patient was found to be normokalemic and normotensive on a double antihypertensive therapy (ACE-inhibitor and Ca blocker). Her most recent visit to our unit was in June 2013. Acromegaly was well controlled on cabergoline 2×0.5 mg weekly. Hormonal parameters showed IGF-1 to be 20.1 nmol/l (normal range: 14–40.5), serum GH during OGTT was: 0 min 2.7 mUI/l; 60 min 1.6 mUI/l; 120 min 1.4 mUI/l (criteria for acromegaly control: basal GH<1 ng/ml (3 mIU/l) and <0.4 ng/ml (1.2 mIU/l) during OGTT11). The last MRI of the pituitary gland, performed in 2010, did not show any enlargement of the existing tumor remnant. During the most recent visit in June 2013 the patient had slightly elevated blood pressure, well controlled using small doses of an ACE- inhibitor (10 mg daily lisinopril) and a Ca-blocker (20 mg daily nifedipine). Laboratory data showed normal potassium of 5.1 mmol/l. The association of acromegaly and Conn’s adenoma raised the hypothesis for MEN1 in our patient, therefore the parathyroid glands were investigated using ultrasound. There were no clinical, laboratory or ultrasound data for primary hyperparathyroidism. Biochemical parameters were normal: serum calcium levels were 2.42 mmol/l (normal range: 2.12–2.62) and phosphorus was 1.35 (normal range: 0.87–1.45). The ultrasound image was not suspicious for parathyroid adenoma. The patient had no relatives with endocrine tumors.\n\n\nDiscussion\n\nWe describe a rare combination of acromegaly and PA due to an aldosterone-producing adenoma. There are several recent reports which describe the association of PA and different types of tumors1–3. These cases are related mainly to MEN1. MEN1 is caused by germline mutations of the menin gene, which most frequently leads to the development of primary hyperparathyroidism, pituitary adenomas and pancreatic tumors. The prevalence of pituitary tumors in MEN1 varies according to different studies, from 10 to 76%12–14. Somatotropinomas causing acromegaly occur in 3–6% of MEN1 patients14,15. As for adrenal involvement in patients with MEN1, the data are contradictory with a prevalence ranging from 9 to 73% depending on the investigated series16–18. A recent large study in 715 MEN1 patients showed adrenal lesions in 146 cases, and among them 72 had adrenal adenomas, comprising 10% of the whole cohort5. Four cases of PA were found in the whole group. None of them was associated with acromegaly, but rather with primary hyperparathyroidism, prolactinoma, or pancreatic endocrine tumors.\n\nIn the literature only single cases of PA have been described in MEN11–3. Beckers et al.1 reported a case of PA as a part of MEN1, associated with parathyroid adenoma, prolactinoma and toxic multinodular goiter. Kim et al.2 described a case of PA, associated with primary hyperparathyroidism, Hurthle cell thyroid cancer and meningioma with a loss of heterozygosity (LOH) of the MEN1 locus in parathyroid glands, but no germline mutation. Honda et al.3 reported PA associated with primary hyperparathyroidism and breast cancer and a LOH of the MEN1 locus in the parathyroid adenoma and in the breast cancer tissue. None of the reported cases was associated with acromegaly, in contrast to our patient. Although investigations for primary hyperparathyroidism were negative in our case, the patient showed two endocrine tumors over several years, which could be related to the presence of MEN1. This could not be confirmed as genetic analysis was not performed due to funding restrictions, which is a serious weak point of the case presentation. However, in different series in up to 60% of patients with a sporadic MEN1 phenotype, no mutation of the corresponding menin gene has been found19.\n\nOn the other hand we could speculate that the activated GH-IGF1 axis may have a role in the morphological and functional adrenal changes in aldosterone-producing adenomas in acromegaly. Although the incidence of neoplasms in acromegaly is a matter of debate, numerous reports have suggested that patients are at increased risk of developing thyroid nodules and colon polyps20, as well as thyroid cancer21 breast cancer22 and colorectal cancer23. There are limited data in the literature concerning the adrenal involvement in acromegalic patients. In a study by Scaroni et al. adrenal morphological abnormalities were found in 28.7% of acromegalic patients (n=94), among them nine cases with unilateral adenoma and the rest with uni- or bilateral hyperplasia6. All tumors were hormonally inactive and no cases of PA were found.\n\nAnother recent study found an even higher prevalence of adrenal abnormalities in patients with acromegaly, describing abnormal adrenal morphology in 48% of patients in a group with 670 acromegalics. Among them 19 patients had an adrenal adenoma, 10 subjects had adrenal hyperplasia and 7% had hyperaldosteronism7. The increased prevalence of adrenal incidentaloma in acromegaly suggests that GH and IGF1 may have an effect on adrenal morphological changes, although in both studies cited above, no significant correlation with GH-IGF1 were found. On the other hand, it has been shown that IGF1 factors are potent mitogens and a strong IGFII expression has been demonstrated in adrenocortical tumors24. Also, a recent study demonstrated that the GH receptor is expressed in both normal rat and normal and diseased human adrenals, which suggests direct action of GH in adrenal tissue25.\n\nFrom a functional point of view several studies have explored the interactions between the GH-IGF1 axis and RAAS. A stimulatory effect of the GH/IGF1 on RAAS has been demonstrated in some reports8,9,26. A recent study by Bielohuby et al.27 showed increased levels of aldosterone in acromegalic patients which normalized after surgery. Aldosterone levels were elevated in a transgenic mouse model over-expressing GH compared to non-transgenic mice and changes in aldosterone were independent of IGF1, renin and the expression of aldosterone synthase27. These studies suggest a direct effect of GH on adrenal glands, which could lead to abnormalities of function or morphology and eventually to the formation of an aldosterone-producing adenoma as may have occurred in our patient.\n\n\nConclusion\n\nWe have described an uncommon case of two endocrine tumors – somatoprolactinoma and Conn’s adenoma diagnosed years after initial presentation of acromegaly. This combination could be part of the MEN1 syndrome, despite the absence of hyperparathyroidism. On the other hand, our patient already had several co-morbidities due to the mitogenic effect of the GH-IGF1 hyperactivation (multinodular goiter, myoma and colon polyp), which is another plausible hypothesis for the functional and structural changes in the adrenal gland.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nJoanna Matrozova and Silvia Vandeva wrote the manuscript, Sabina Zacharieva revised the manuscript. All the authors approved the final manuscript for publication.\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\nBeckers A, Abs R, Willems PJ, et al.: Aldosterone-secreting adrenal adenoma as part of multiple endocrine neoplasia type 1 (MEN1): loss of heterozygosity for polymorphic chromosome 11 deoxyribonucleic acid markers, including the MEN1 locus. J Clin Endocrinol Metab. 1992; 75(2): 564–570. PubMed Abstract \| Publisher Full Text\n\nKim YL, Jang YW, Kim JT, et al.: A rare case of primary hyperparathyroidism associated with primary aldosteronism, Hurthle cell thyroid cancer and meningioma. J Korean Med Sci. 2012; 27(5): 560–564. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHonda M, Tsukada T, Horiuchi T, et al.: Primary hyperparathyroidism associatiated with aldosterone-producing adrenocortical adenoma and breast cancer: relation to MEN1 gene. Intern Med. 2004; 43(4): 310–314. PubMed Abstract \| Publisher Full Text\n\nBrandi ML, Gagel RF, Angeli A, et al.: Guidelines for diagnosis and therapy of MEN type 1 and type 2. J Clin Endocrinol Metab. 2001; 86(12): 5658–5671. PubMed Abstract \| Publisher Full Text\n\nGatta-Cherifi B, Chabre O, Murat A, et al.: Adrenal involvement in MEN1. Analysis of 715 cases from the Groupe d'etude des Tumeurs Endocrines database. Eur J Endocrinol. 2012; 166(2): 269–279. PubMed Abstract \| Publisher Full Text\n\nScaroni C, Selice R, Benedini S, et al.: Adrenal morpho-functional alterations in patients with acromegaly. J Endocrinol Invest. 2008; 31(7): 602–606. PubMed Abstract\n\nPappa T, Papanastasiou L, Koutmos S, et al.: Pattern of adrenal morphology and function in patients with acromegaly. Eur J Clin Invest. 2012; 42(3): 275–281. PubMed Abstract \| Publisher Full Text\n\nCiresi A, Amato MC, Vetro C, et al.: Adrenal morphology and function in acromegalic patients in relation to disease activity. Endocrine. 2009; 36(2): 346–354. PubMed Abstract \| Publisher Full Text\n\nMantero F, Opocher G, Armanini D, et al.: Plasma renin activity and urinary aldosterone in acromegaly. J Endocrinol Invest. 1979; 2(1): 13–18. PubMed Abstract\n\nStrauch G, Vallotton MB, Touitou Y, et al.: The renin-aldosterone system in acromegaly: follow-up report. N Engl J Med. 1973; 289(15): 808. PubMed Abstract\n\nGiustina A, Chanson P, Bronstein MD, et al.: A consensus on criteria for cure of acromegaly. J Clin Endocrinol Metab. 2010; 95(7): 3141–3148. PubMed Abstract \| Publisher Full Text\n\nLourenco DM Jr, Toledo RA, Mackowiak II, et al.: Multiple endocrine neoplasia type 1 in Brazil: MEN1 founding mutation, clinical features, and bone mineral density profile. Eur J Endocrinol. 2008; 159(3): 259–274. PubMed Abstract \| Publisher Full Text\n\nBurgess JR, Shepherd JJ, Parameswaran V, et al.: Prolactinomas in a large kindred with multiple endocrine neoplasia type 1: clinical features and inheritance pattern. J Clin Endocrinol Metab. 1996; 81(5): 1841–1845. PubMed Abstract \| Publisher Full Text\n\nMarx S, Spiegel AM, Skarulis MC, et al.: Multiple endocrine neoplasia type 1: clinical and genetic topics. Ann Intern Med. 1998; 129(6): 484–494. PubMed Abstract \| Publisher Full Text\n\nTrump D, Farren B, Wooding C, et al.: Clinical studies of multiple endocrine neoplasia type 1 (MEN1). QJM. 1996; 89(9): 653–669. PubMed Abstract \| Publisher Full Text\n\nWaldmann J, Bartsch DK, Kann PH: Adrenal involvement in multiple endocrine neoplasia type 1: results of 7 years prospective screening. Langenbecks Arch Surg. 2007; 392(4): 437–443. PubMed Abstract \| Publisher Full Text\n\nSchaefer S, Shipotko M, Meyer S, et al.: Natural course of small adrenal lesions in multiple endocrine neoplasia type 1: an endoscopic ultrasound imaging study. Eur J Endocrinol. 2008; 158(5): 699–704. PubMed Abstract \| Publisher Full Text\n\nSkogseid B, Larsson C, Lindgren PG, et al.: Clinical and genetic features of adrenocortical lesions in multiple endocrine neoplasia type 1. J Clin Endocrinol Metab. 1992; 75(1): 76–81. PubMed Abstract \| Publisher Full Text\n\nHai N, Aoki N, Shlmatsu A, et al.: Clinical features of multiple endocrine neoplasia type 1 (MEN1) phenocopy without germline MEN1 gene mutations: analysis of 20 Japanese sporadic casis with MEN1. Clin Endocrinol (Oxf). 2000; 52(4): 509–518. PubMed Abstract \| Publisher Full Text\n\nColao A, Ferone D, Marzullo P, et al.: Systemic complications of acromegaly: epidemiology, pathogenesis, and management. Endocr Rev. 2004; 25(1): 102–152. PubMed Abstract \| Publisher Full Text\n\nTita P, Ambrosio MR, Scollo C, et al.: High prevalence of differentiated thyroid carcinoma in acromegaly. Clin Endocrinol (Oxf). 2005; 63(2): 161–167. PubMed Abstract \| Publisher Full Text\n\nCheung NW, Boyages SC: Increased incidence of neoplasia in females with acromegaly. Clin Endocrinol (Oxf). 1997; 47(3): 323–327. PubMed Abstract \| Publisher Full Text\n\nRokkas T, Pistiolas D, Sechopoulos P, et al.: Risk of colorectal neoplasm in patients with acromegaly: a meta-analysis. World J Gastroenterol. 2008; 14(22): 3484–3489. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFottner C, Hoeflich A, Wolf E, et al.: Role of the insulin-like growth factor system in adrenocortical growth control and carcinogenesis. Horm Metab Res. 2004; 36(6): 397–405. PubMed Abstract \| Publisher Full Text\n\nLin CJ, Mendonca BB, Lucon AM, et al.: Growth hormone receptor messenger ribonucleic acid in normal and pathologic human adrenocortical tissues--an analysis by quantitative polymerase chain reaction technique. J Clin Endocrinol Metab. 1997; 82(8): 2671–2676. PubMed Abstract \| Publisher Full Text\n\nNatchev E, Kirilov G, Atanasova I, et al.: Changes of some humoral factors playing role in the pathogenesis of hypertension in acromegaly. Endocrinolgia. 2011; 16: 177–189.\n\nBielohuby M, Roemmler J, Manolopoulou J, et al.: Chronic growth hormone excess is associated with increased aldosterone: a study in patients with acromegaly and in growth hormone transgenic mice. Exp Biol Med (Maywood). 2009; 234(8): 1002–1009. PubMed Abstract \| Publisher Full Text" }	[ { "id": "3909", "date": "10 Mar 2014", "name": "Niki Karavitaki", "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 of a patient with co-existing acromegaly and primary aldosteronism attributed to a Conn's adenoma.Although the biochemical confirmation of primary aldosteronism relied only on a Captopril test and no adrenal vein sampling took place, the improvement of the blood pressure control and the resolution of hypokalaemia following the removal of the adrenal mass support the diagnosis of Conn's adenoma.The authors attempt to provide a pathophysiological link between the two conditions. Their thoughts are very interesting but the option of a mere coincidence also needs to be mentioned.", "responses": [ { "c_id": "831", "date": "22 May 2014", "name": "Silvia Vandeva", "role": "Author Response", "response": "Dear Dr Karavitaki,The authors thank you for your comments. We have conformed to your recommendations." } ] }, { "id": "4554", "date": "25 Apr 2014", "name": "Shlomo Melmed", "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 described case of interest to readers. It is clearly written and contains important information.", "responses": [ { "c_id": "830", "date": "22 May 2014", "name": "Silvia Vandeva", "role": "Author Response", "response": "Dear prof. Melmed,The authors thank you for your comments." } ] } ]	1	https://f1000research.com/articles/3-58
https://f1000research.com/articles/2-266/v1	05 Dec 13	{ "type": "Review", "title": "Periprosthetic bone loss: diagnostic and therapeutic approaches", "authors": [ "Loredana Cavalli", "Maria Luisa Brandi", "Loredana Cavalli" ], "abstract": "Total joint replacement surgery is being performed on an increasingly large part of the population. Clinical longevity of implants depends on their osseointegration, which is influenced by the load, the characteristics of the implant and the bone-implant interface, as well as by the quality and quantity of the surrounding bone. Aseptic loosening due to periprosthetic osteolysis is the most frequent known cause of implant failure. Wear of prosthetic materials results in the formation of numerous particles of debris that cause a complex biological response. Dual-energy X-ray Absorptiometry (DXA) is regarded as an accurate method to evaluate Bone Mineral Density (BMD) around hip or knee prostheses. Further data may be provided by a new device, the Bone Microarchitecture Analysis (BMA), which combines bone microarchitecture quantification and ultra high resolution osteo-articular imaging. Pharmacological strategies have been developed to prevent bone mass loss and to extend implant survival. Numerous trials with bisphosphonates show a protective effect on periprosthetic bone mass, up to 72 months after arthroplasty. Strontium ranelate has been demonstrated to increase the osseointegration of titanium implants in treated animals with improvement of bone microarchitecture and bone biomaterial properties.", "keywords": [ "periprosthetic bone loss", "osteolysis", "stress shielding", "subsidence", "BMD", "DXA", "BMA", "implant", "arthroplasty", "bisphosphonate", "strontium ranelate" ], "content": "Introduction\n\nEndosseous implantation is one of the most common procedures in orthopedics and dentistry. The ever expanding aging population has also led to an increasing need for total joint replacements1. Unfortunately, the introduction of a prosthesis or a dental implant inevitably alters the physiological transmission of loads to the surrounding bone, which starts a remodeling process, resulting in reduction in bone mineral density (BMD)2.\n\nAseptic loosening due to bone destruction around the prosthesis has been established as the main cause of implant failure3–6. Mechanical, thermal and chemical intraoperative damage induces necrotic phenomena on the periprosthetic bone, which takes approximately 3 months to repair2,7. Then the osteocytes, acting as mechanoreceptors, translate the mechanical stimulus into an electrical signal, either activating osteoclasts (OCs) in bone areas no longer subjected to physiological loading or stimulating osteoblastic cell lines where bone is stressed, with consequent hypertrophy2,8–10.\n\nDespite being a widespread practice, joint arthroplasty almost unavoidably involves a loss of surrounding bone, which can cause periprosthetic fractures resulting in reduced function, subsequent morbidity and increased risk of mortality11. After Total Hip Arthroplasty (THA), for example, periprosthetic fractures occur in 0.8% of patients at 5 years, and 3.5% at 10 years11,12. This is the third most common reason for re-operation13,14, while implant failure by aseptic loosening could be expected in 3–10% of cases within 15 years11,12.\n\nA better understanding of the biological basis of peri-implant osteolysis has allowed the development of therapeutic strategies to prevent periprosthetic bone loss, in particular with Bisphosphonates (BPs) and Strontium Ranelate (SrRan).\n\nThis paper will cover the biological basis of periprosthetic bone resorption, diagnostic techniques and preventive or therapeutic approaches, both from a pharmacological and surgical point of view.\n\n\nPathogenic mechanisms underlying aseptic loosening of implants\n\nRisk factors for periprosthetic bone destruction include osteoporosis, rheumatoid arthritis, revision surgery and stress shielding. These lead to a resorption process in bone areas that are no longer mechanically subjected15. The basis of this process includes mechanical and biological factors16.\n\nSeveral reports have shown that the cellular responses to biomaterial wear particles play an important role16. Particles ranging from 0.2 to 10 μm in diameter undergo phagocytosis by macrophages17.\n\nIn vitro studies of macrophage cultures clearly indicate that smaller particles of polymethylmethacrylate (PMMA) and polyethylene, materials used in implants, (< 20 μm) elicit a significantly increased inflammatory cytokine response, as indicated by increased release of Tumor Necrosis Factor (TNF), IL-1, IL-6, prostaglandin (PG)E2, matrix metalloproteinases, receptor activator of nuclear factor kappa-B ligand (RANKL) and other factors that affect osteoclast differentiation and activity16,18–21. Moreover, direct biologic interaction between particles and the cell surface is sufficient to activate osteoclastogenic signaling pathways16,18,22, causing bone resorption and periprosthetic BMD loss.\n\nBone turnover markers, due to their high intra-individual variability, have a limited predictive value concerning the extent of periprosthetic bone loss23. However, the immediate high postoperative activity of osteoclasts is confirmed by a study on 53 patients followed for 12 months after THA with a cemented femoral stem23. The study showed an early postoperative increase of C-terminal telopeptides of type I collagen (CTX-I) (markers of bone resorption), where the highest level was noted (+21%) 3 weeks after THA, then decreased at 8 weeks after THA (-7% from preoperative levels). This increase was significantly correlated with the bone loss measured by DXA in the calcar region23. These data suggest that a postoperative antiresorptive treatment administered for the period of increased CTX-I levels could prevent periprosthetic bone loss23.\n\nStress shielding is also considered as a potent stimulator of bone resorption. After a total hip arthroplasty, the stem geometry of the implant plays a key role in the load transfer to the femur and consequently in femoral remodeling24,25.\n\nConflicting results have emerged regarding the hypothesis that the amount of periprosthetic bone remodeling can be influenced by different factors, including sex, age, weight, Body Mass Index (BMI) and bone mass26–30. The data on the role of stem design are more consistent24.\n\n\nImaging techniques for the evaluation of periprosthetic BMD loss\n\nDual-energy X-ray Absorptiometry (DXA) is considered the most reliable tool to evaluate bone remodeling after THA using implants with different stem designs31,32. It is also used to assess the effectiveness of these treatments by comparing the medium-term bone density changes between treated and untreated groups. A special piece of software named “metal-removal” enables DXA to analyze periprosthetic bone using seven conventional Regions Of Interest (ROI) called Gruen zones (Figure 1)33,34.\n\nThanks to improvements in software and technology, bone densitometry examinations by DXA may actually allow the detection of periprosthetic bone remodeling that cannot be observed in conventional radiographs as DXA provides an accurate measurement of total and regional periprosthetic BMD after THA24,35–38.\n\nDXA scanning is usually performed with the patient in the supine position, the leg placed in a standardized support to ensure a neutral position38. Analysis of the 7 periprosthetic Gruen zones is the most commonly used protocol to evaluate bone remodeling after the implantation of conventional femoral stems24,31,40,41. As shown in Figure 1, in the horizontal plane, the tip of the lesser trochanter defines the distal border of zones 1 and 7. The midpoint between the lesser trochanter and the tip of the stem defines the border between zone 2 and zones 3, 5 and 6. Zone 4 represents the total bone area 20 mm distally from the tip of the stem. Vertically, the center axis of the femur divides the medial and lateral zones39.\n\nPostoperative measurements are commonly used as baseline values and the measurements at follow-up are expressed as a percentage of the baseline measurements39,42. However, in cross-sectional studies, the controlateral unoperated hip has also been analyzed to obtain individual comparative BMD values24,40,43,44.\n\nAlthough DXA is regarded as the most accurate method for the detection of small alterations in bone mineral density around hip prostheses39, its metal-removal software also provides peri-prosthetic measurements for knee arthroplasty30,45–49. No DXA protocol is available for ankles, shoulders, elbows, or wrists.\n\nAn innovative device has recently been developed, the BMA, which combines ultrahigh resolution 2D digital X-Ray images and a set of trabecular bone texture analysis parameters, such as 2D fractal analysis (H mean), Co-Occurrence (COOC) and Run Length Encoding (RLE), thus providing a bone microarchitecture quantification independent of bone density measurement50.\n\nWith a resolution near 100 µm, BMA visualizes the bone structure at the trabecular level, allowing the in vivo micro-analysis of human bone structure and abnormalities, such as fracture lines that are often uncertain or ignored. All the joints can be examined (i.e. spine, hip, knee, ankle, shoulder, elbow, wrist and phalanges). The micrometric accuracy in the visualization of joint interspace reflects cartilage thickness, useful for the diagnosis and follow-up of osteoarthritis (OA). The digital X-Ray detector allows examination at a very low dose (effective dose < 2 μSV for a heel exam) due to its excellent Detection Quantum Efficiency (DQE). This performance, added to the quality of a high frequency X-Ray generator, is obtained in less than 1 second exposure time, while the image processing is achieved in less than 2 minutes, facilitating patient workflow and improving productivity50.\n\nThe employment of BMA may represent a promising device for the study of periprosthetic bone analysis of any joint as well as for arthritis, osteoarthritis and altered bone healing.\n\n\nTherapeutic strategies for enhancing bone mass recovery after arthroplasty\n\nSince the stem geometry of the implant is believed to play an important role in load transfer to the femur, biomechanical tests51–53 and radiographic studies54, followed by DXA analyses, have been conducted on patients subjected to THA. These studies suggest that the ultra-short implant (which has a more anatomical proximal fit without having a diaphyseal stem with distal cortical contact) can provide immediate postoperative stability and a more physiological load distribution, thus increasing periprosthetic BMD in the medial regions over time24, preserving bone mass and stimulating trabecular bone apposition24,53,55. The presence of the lateral flare makes the diaphyseal stem with distal cortical contact almost unnecessary, thus increasing periprosthetic BMD in the medial regions over time24.\n\nAlbanese and colleagues24 assessed bone remodeling in patients subjected to two metaphyseal implants, type 1 with a very short stem and type 2 with no diaphyseal stem. Using a 5-ROI protocol of DXA analysis, they found that ultra-short implants can provide immediate postoperative stability and a more natural physiological load distribution in comparison with conventional anatomic implants, thus increasing periprosthetic BMD in the medial regions over time.\n\nAfter stem design, the most important factor known to influence periprosthetic BMD is the fixation of the implant27. A fundamental feature that enables fixation is the porous surface of the prosthesis56. Cementless THA is increasingly popular. The high rate of osteolysis, aseptic loosening and revision associated with earlier uncemented femoral components has been greatly reduced by using better designed implants incorporating circumferential porous coating56. Moreover, proximal femoral fixation has been shown to prevent stress shielding and a tapered distal tip reduces thigh pain56. The mid-term outcome of a modular, cementless, proximally hydroxyapatite-coated, anatomic femoral stem in THA was reported by Cossetto and Goudar (Modulaire Biconique Anatomique, MBA Groupe Lépine, France)56. They showed that the modularity of the neck of this femoral component is helpful in both primary and revision settings. In primary procedures, after implantation of the stem, correction can still be made in leg length and offset. In revision procedures, the modularity of the neck facilitates adjustments in leg length, offset and neck version without the need to extract a well fixed femoral component. It also facilitates access to the acetabular component by way of removal of the modular neck and head56. In that study, in case of dislocation, changing the modular neck and head avoided more extensive revision requiring removal of a well-fixed femoral stem. Patients were evaluated pre- and post-operatively (at 6 weeks, 3 months, one year, 2 years, 5 years and 10 years), with a clinical evaluation (pain, range of movement, and ability to walk) based on Merle d’Aubigne and Postel scores57 and by anteroposterior and lateral weight-bearing radiographs, in which the femoral component was analyzed according to the 7 zones of Gruen. Contrary to the increased rate of revision in modular hip systems found in the Australian Orthopaedic Association National Joint Replacement Registry, the modular, cementless, proximally hydroxyapatite-coated, anatomic femoral stem provided predictably stable fixation with excellent mid-term outcome58.\n\nSimilarly, Lerch and colleagues33 conducted a prospective densitometric study by DXA in a group of patients who underwent unilateral bicontact stem implantation. This is a cementless implant made of a titanium forged alloy (Ti6A14V), with a proximal microporous pure titanium plasmapore coating. Despite small signs of stress shielding observed at the tip of the stem, it has shown to provide adequate proximal bone stock preservation33.\n\nStudying knee arthroplasty is rather difficult when compared to studying THA, due to the position of the patient required for the exam. Full extension of the knee is not possible for most patients in the first days after surgery, therefore while deficits in extension normalize with rehabilitation, individual knee flexion between the baseline and follow-up investigations may be different47. A clinically applicable soft foam positioner designed to ensure rotational stability and allow for slight flexion (i.e. 25°) may be safe for clinical use, because this position can be obtained with all normal total knee arthroplasty (TKA) patients both in the early period after surgery and in later follow-ups47.\n\nHowever, a prospective cohort study conducted by Windisch and colleagues48 described the changes in bone density over the course of time following a cement-free TKA based on a functional categorization of the measurements in terms of defined ROI by means of DXA. The seven regions were defined as indicated in Table 1.\n\nA further aim of that study was to examine the associations between the defined parameters of age, sex, severity of arthrosis, and axis alignment. At 12 months after surgery, a high severity of osteoporosis was associated with low absolute values for periprosthetic bone density. Women demonstrated a lower absolute periprosthetic BMD value than men. The preoperatively determined femur and tibial average Cortical bone Marrow Index (CMI), the varus angle and the BMI showed no significant correlation with the absolute or relative changes of periprosthetic bone density48. Statistical analysis revealed that the most significant changes occurred within the first 3 months postoperative with the highest bone density loss found in the region of the proximal medial tibia48.\n\nAnother group49 has studied the effects of Unicompartmental Knee Arthroplasty (UKA), which has received renewed interest for medial OA within the last decade. UKA has been traditionally used in older, non-obese patients with a sedentary lifestyle. UKA is advantageous as only the severely damaged compartment is replaced and the bone stock is preserved, which is associated with fast recovery times49. Moreover, improvements in surgical technique, implant materials and prosthetic design have made UKAs more durable and reliable49.\n\nThe group measured BMD using DXA and data were collected from multiple ROIs for each patient at several intervals during the first 7 postoperative days. The highest femoral periprosthetic bone loss rate was observed during the first 3 months after UKA. However, BMD changes from 2–7 years were not significant. In particular, there was a significant loss of BMD from distal femoral sites after UKA, while BMD changes were minor in the tibial metaphyseal regions, consistent with a mechanical axis balance between the medial and lateral sides of the tibia. Further, porous tantalum tibial components maintained better periprosthetic BMD compared with cemented tibial implants49,59.\n\nIn combination with improvements in implant integration, strategies to target the cellular components (osteoblasts and osteoclasts) that contribute to implant failure should be implemented60. In this regard, it should be noted that differentiation of bone marrow macrophages into mature osteoclasts requires recognition and binding of osteoblasts, fibroblasts, and T cell secreted factor RANKL by its cognate receptor RANK, which is expressed on the surface of osteoclast precursors61–63. Another osteoblastic factor, namely osteoprotegerin (OPG), acts as a decoy receptor by binding to RANKL and reducing its bioavailability. On the other hand, binding of RANKL to RANK stimulates induction of several intracellular pathways by this receptor, leading to activation of key transcription factors, most notably NF-κB22.\n\nIt is known that NF-κB activation, when induced by factors such as TNF and PMMA particles, exacerbates osteoclastogenesis and inflammatory responses22. In this context, a review by the group of Abu-Amer considers three kinds of approaches22. The first involves targeting OC precursor cells, which are brought to inflammatory sites by circulating cytokines. The second entails targeting precursors that are stimulated by the particle-mediated cellular response to differentiate into OCs. The third approach involves targeting activation mechanisms of mature osteoclasts22.\n\nAn example of the first strategy is the application of RANKL decoy molecules such as OPG and the soluble fusion protein RANK-Fc64,65. At present, the monoclonal antibody anti-RANKL is available, known as ‘denosumab’, which shows significant effectiveness in the inhibition of bone resorption due to osteoporosis66. Transduction of a dominant-negative form of the NF-κB inhibitory protein, IκB, by retaining NF-κB in the cytoplasm, has been revealed to be able to block osteoclast formation and activity16,67–69. Another viable approach is to block activation of the upstream IKK complex, which is responsible for phosphorylation of IκB and subsequent activation of NF-κB, by introducing a small peptide that hinders assembly of the IKK complex16,70. Notably, administration of the dominant negative IκB protein or the IKK inhibitory small peptide to arthritic mice blocks bone erosion and particle-induced osteolysis of calvaria in mice16,71.\n\nFurther targets of therapy are NF-κB mediated genes. Recent studies have revealed that proinflammatory cytokines such as TNF act directly on some of these genes and their products, in particular c-src and NF-κB, to accelerate osteoclast formation and cause a strong osteoclastic response71. Selective inhibitors of the c-src tyrosine kinase have shown great promise in reducing osteoclast activity16,72,73.\n\nAnother promising approach involves the use of bisphosphonates (BP)2,16,74, potent anti-resorptive drugs widely used in the treatment of osteoporosis, which inhibit osteoclast function and induce their apoptosis. In animal models subjected to implantation75–77, oral BP showed reduced radiographic periprosthetic radiolucency, as inhibiting debris-induced osteolysis, although the levels of PGE2 and IL-1 remained elevated in tissue cultures from these implants. These studies have served as a basis for clinical trials using alendronate, one of the most commonly used BPs, in patients with radiographically evident osteolytic lesions. In other studies, bone loss around implants was prevented and treated by alendronate78–81.\n\nMore recently, human clinical trials have revealed the efficacy of BPs in reducing particle-induced osteolysis over the first year of life of cemented and cementless hip and knee replacement prostheses, with better and more durable results when treatment was started early after surgery and continued for over 6 months2,82,83. Moreover, in vivo trials showed a direct action of some BPs in stimulating the osteoblastic proliferation, which might play an essential role in increasing periprosthetic bone ingrowth84. The mechanism by which BPs are supposed to act on the osteoblasts is by up-regulating the expression of genes coding the synthesis of some morphogenetic proteins, including BMP-285.\n\nMuratore and colleagues assessed the effect of ibandronate, another type of BP that can peculiarly be administered either orally or intravenously, with extended dosing intervals, thanks to its high affinity for the bone mineral component and its consequent long-term persistence in the skeletal tissue, which therefore ensures excellent adherence to therapy2. Thirty-five women over 60 years old, not necessarily suffering from osteoporosis, were subjected to THA. They were examined by DXA at the 15th day after surgery (T0) and at 6 and 12 months, either at the spine, contralateral femur or periprosthetic femur, both totally and at the 7 Gruen regions. Of these patients, 19 patients received 3 mg ibandronate intravenously within 5 days after surgery and then passed to oral administration with a monthly dose of 150 mg, plus calcium carbonate (1 g) and cholecalciferol (880 IU) supplementation. The other 16 patients formed the control group and were treated only with calcium carbonate (1 g) and cholecalciferol (880 IU) supplementation. As a result, a reduction in the BMD was observed over the first 6 months from T0 in both groups; smaller reductions were observed in the treated group (-7.7% compared to the control group). In contrast, at 12 months, a marked trend reversal was observed, with a statistically significant BMD percentage recovery compared to the baseline value at T0 of about 1.74% of the global BMD in the treatment group. This was more evident in region R1 (+3.81%) and in the lateral metaphyseal region (R2) (+4.12%). On the other hand, no global BMD recovery was observed in the control group, which had virtually stabilized compared to values at 6 months2.\n\nConsidering that periprosthetic remodeling occurs within the first 6-12 months after surgery, this study therefore concluded that ibandronate reduces periprosthetic resorption, in particular in the medial metaphyseal region (calcar and lesser trochanter), the one at greater risk with respect to the life of the prosthesis.\n\nBisphosphonates, which are chemically stable analogues of inorganic pyrophosphate, can be classified into at least two groups with different molecular modes of action. The simpler non-nitrogen-containing bisphosphonates (such as etidronate and clodronate, BPs of first generation) can be metabolically incorporated into nonhydrolysable analogues of adenosine triphosphate, which interfere with adenosine triphosphate-dependent intracellular pathways86. The more potent nitrogen-containing bisphosphonates (including pamidronate, alendronate, risedronate, ibandronate, and zoledronate) are not metabolized in this way but inhibit key enzymes of the mevalonate/cholesterol biosynthetic pathway, such as farnesyl pyrophosphate synthase, compromising the function of essential intracellular messengers, thus causing osteoclast inactivation and apoptosis87.\n\nA meta-analysis11 of 14 randomized controlled trials employing BPs after joint arthroplasty found that the protective effect of these drugs, probably modified by BP generation and the prosthesis location, could persist in a middle-term follow-up after surgery and for 18 to 70 months after drug discontinuation. The efficacy was more potent for amino-BPs, than for the first generation of BPs11.\n\nHowever, since the trials did not address the clinically relevant outcomes, it is imperative to perform a randomized clinical trial with an adequate number of patients and sound methodology in order to establish the definitive role of BPs in joint arthroplasty, and make recommendations for their optimal administration. It is also necessary to better understand the mechanisms of their actions and potential side effects11.\n\nMoreover, prolonged use of BPs has recently been associated with severe suppression of bone turnover88–90, alterations in normal collagen cross-linking and matrix heterogeneity91–93, reduced vascularity94,95 and decreased cortical bone toughness95–98, as well as a small number of subtrochanteric or diaphyseal femoral fractures95,99–102. The exact mechanism of these ‘atypical’ fractures is unknown. One theory is that they occur in the subtrochanteric region of the femur because it is subject to high bending forces103–106. The latter would cause the formation of micro-cracks, normally repaired through bone remodeling. Bisphosphonates’ suppression of bone turnover results in a failure to repair these micro-cracks102.\n\nCalcitonin is a 32-amino acid polypeptide hormone (produced predominantly in C-cells of the thyroid gland) which mainly acts by inhibiting osteoclast function107. It has been shown that 200 IU of salmon calcitonin administered nasally decreases osteoporotic fractures108,109. A Finnish clinical trial107 randomized 60 patients who underwent THA using cemented Exeter prostheses into a treatment group (200 IU salmon calcitonin + calcium 500 mg) and placebo group (inactive nasal spray + calcium 500 mg) for 6 months. They were followed with DXA, bone turnover markers and dynamic histomorphometry on bone biopsies taken from the femoral neck at the time of discharge, after 6 and 12 months. Calcitonin was not shown to promote any additional value on calcium substitution in preventing aseptic osteolysis107.\n\nAnother antiosteoporotic treatment, Strontium Ranelate (SrRan), was shown to be of considerable interest in investigations to improve implant osseointegration1. The beneficial effects of SrRan have previously been reported in various animal models, where it has been shown to prevent bone loss by maintaining bone formation at a high level and inhibiting bone resorption1,110–114. These in vivo results were consistent with in vitro data which show that SrRan reduces bone resorption by osteoclasts and increases bone formation by osteoblasts115–117. It has also been demonstrated that SrRan is able to improve bone biomechanical and structural properties118. Furthermore, treatment with SrRan is not associated either with osteonecrosis of the jaw or with low energy atypical fractures of the femur1.\n\nA group from the Bone Division of Geneva1 showed that SrRan significantly improves mechanical fixation of titanium implants inserted into the tibias of female rats, with both a positive effect on bone microarchitecture and on bone biomaterial properties in the vicinity of the implant. SrRan increased pull-out strength compared to controls (+34%), with a significant improvement of bone microarchitecture around the implant, a more plate-shape structure and an increase in bone-to-implant contact (+19%)1.\n\nAnother study119 was designed to evaluate the effect of systemic SrRan treatment on fixation of hydroxyapatite (HA)-coated titanium screws in ovariectomized (OVX) rats. The OVX rats were randomly divided into the following groups: OVX (without treatment), OVX+SRL (treated with a low SR (i.e. SrRan) dose of 500 mg/kg/day) and OVX+SRH (with a high SR dose of 1000 mg/kg/day). Micro-CT and biomechanical push-out tests were performed twelve weeks after treatment, in order to evaluate bone blocks with implants. The two groups treated with SR showed an increase of bone volume ratio, osseointegration and maximal force, compared to OVX animals, suggesting that SrRan treatment can improve HA-coated screw fixation dose-dependently in OVX rats and facilitate the stability of the implant in the osteoporotic bone119.\n\nThese results may support the potential benefits of SrRan in enhancing osseointegration in orthopaedic and dental surgery.\n\n\nConclusions\n\nAseptic loosening, due to periprosthetic osteolysis, is the most common cause of implant failure.\n\nAmong the other clinical and biomechanical criteria, bone status should be considered before proceeding with arthroplasty, in order to select the most adequate implant model as well as to evaluate the necessity of an anti-osteoporosis therapy.\n\nThe comprehension of the principal mechanisms of periprosthetic bone loss has led to the development of pharmacologic strategies aiming at the enhancement of bone mass recovery after surgery and consequently to the prolongation of implant survival.\n\nBPs, potent anti-resorptive drugs widely used in the treatment of osteoporosis and other disorders of bone metabolism, were shown to be particularly effective in reducing periprosthetic bone resorption in the first year after hip and knee arthroplasty, both cemented and cementless.\n\nSrRan, due to its antiresorbing and bone-forming activity, also promises to facilitate the stability of dental and joint implants in both healthy and osteoporotic bone.", "appendix": "Author contributions\n\n\n\nLC prepared the first draft of the manuscript. LC and MLB were involved in the revision 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\nAcknowledgments\n\nThe authors are grateful to Marco Postiglione MD, for the English language revision of the paper.\n\n\nReferences\n\nMaimoun L, Brennan TC, Badoud I, et al.: Strontium Ranelate improves implant osseointegration. Bone. 2010; 46(5): 1436–41. PubMed Abstract \| Publisher Full Text\n\nMuratore M, Quarta E, Quarta L, et al.: Ibandronate and cementless total hip arthroplasty: densitometric measurement of periprosthetic bone mass and new therapeutic approach to the prevention of aseptic loosening. Clin Cases Miner Bone Metab. 2012; 9(1): 50–55. PubMed Abstract \| Free Full Text\n\nGoldring SR, Schiller AL, Roilke M, et al.: The synovial-like membrane at the bone-cement interface in loose total hip replacements and its proposed role in bone lysis. J Bone Joint Surg Am. 1983; 65(5): 575–84. PubMed Abstract\n\nHarris WH, Schiller AL, Scholler JM, et al.: Extensive localized bone resorption in the femur following total hip replacement. J Bone Joint Surg Am. 1976; 58(5): 612–8. PubMed Abstract\n\nJasty MJ, Floyd WE 3d, Schiller AL, et al.: Localized osteolysis in stable non-septic total hip replacement. J Bone Joint Surg Am. 1986; 68(6): 912–9. PubMed Abstract\n\nBoynton E, Waddell JP, Morton J, et al.: Aseptic loosening in total hip implants: the role of polyethylene wear debris. Can J Surg. 1991; 34(6): 599–605. PubMed Abstract\n\nMassari L, De Rito G, Brunoro A, et al.: Il Disodio Clodronato riduce il riassorbimento periprotesico in impianti non cementati nelle fasi precoci: studio prospettico con controllo. GIOT. 2002; 28: 79–87. Reference Source\n\nIolascon G: Bisfosfonati e protesi articolari. Up-to-date 2005. In: Adami S (Ed.) Bisfosfonati Yearbook. 2005; 71–78. Reference Source\n\nMarotti G, Muglia MA, Palumbo C: Collagen texture and osteocyte distribution in lamellar bone. Ital J Anat Embryol. 1995; 100(Suppl. 1): 95–102. PubMed Abstract\n\nRubinacci A, Villa I, Dondi Benelli F, et al.: Osteocyte-bone lining cell system at the origin of steady ionic current in damaged amphibian bone. Calcif Tissue Int. 1998; 63(4): 331–339. PubMed Abstract \| Publisher Full Text\n\nLin T, Yan SG, Cai XZ, et al.: Bisphosphonates for periprosthetic bone loss after joint arthroplasty: a meta-analysis of 14 randomized controlled trials. Osteoporos Int. 2012; 23(6): 1823–1834. PubMed Abstract \| Publisher Full Text\n\nFurnes O, Lie SA, Espehaug B, et al.: Hip disease and the prognosis of total hip replacements. A review of 53,698 primary total hip replacements reported to the Norwegian Arthroplasty Register 1987–99. J Bone Joint Surg Br. 2001; 83(4): 579–586. PubMed Abstract \| Publisher Full Text\n\nMaury AC, Pressman A, Cayen B, et al.: Proximal femoral allograft treatment of Vancouver type-B3 periprosthetic femoral fractures after total hip arthroplasty. J Bone Joint Surg Am. 2006; 88(5): 953–958. PubMed Abstract \| Publisher Full Text\n\nMcLean AL, Patton JT, Moran M: Femoral replacement for salvage of periprosthetic fracture around a total hip replacement. Injury: Int J Care Injured 2012; 43(7): 1166–1169. PubMed Abstract \| Publisher Full Text\n\nDemos HA, Briones MS, White PH, et al.: A Biomechanical Comparison of Periprosthetic Femoral Fracture Fixation in Normal and Osteoporotic Cadaveric Bone. J Arthroplasty. 2012; 27(5): 783–738. PubMed Abstract \| Publisher Full Text\n\nAbu-Amer Y, Darwech I, Clohisy JC: Aseptic loosening of total joint replacements: mechanisms underlying osteolysis and potential therapies. Arthritis Res Ther. 2007; 9(Suppl 1): S6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGelb H, Schumacher HR, Cuckler J, et al.: In vivo inflammatory response to polymethylmethacrylate particulate debris: effect of size, morphology, and surface area. J Ortho Res. 1994; 12(1): 83–92. PubMed Abstract \| Publisher Full Text\n\nGonzalez O, Smith RL, Goodman SB: Effect of size, concentration, surface area, and volume of polymethylmethacrylate particles on human macrophages in vitro. J Biomed Mater Res. 1996; 30(4): 463–473. PubMed Abstract \| Publisher Full Text\n\nShanbhag AS, Jacobs JJ, Glant TT, et al.: Composition and morphology of wear debris in failed uncemented total hip replacement. J Bone Joint Surg Br. 1994; 76(1): 60–67. PubMed Abstract\n\nAbbas S, Clohisy JC, Abu-Amer Y: Mitogen-activated protein (MAP) kinases mediate PMMA-induction of osteoclasts. J Orthop Res. 2003; 21(6): 1041–1048. PubMed Abstract \| Publisher Full Text\n\nO’Keefe RJ, Rosier RN, Teot LA, et al.: Cytokine and matrix metalloproteinase expression in pigmented villonodular synovitis may mediate bone and cartilage destruction. Iowa Orthop J. 1998; 18: 26–34. PubMed Abstract \| Free Full Text\n\nGallo J, Kamínek P, Tichá V, et al.: Particle disease. A comprehensive theory of periprosthetic osteolysis: a review. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2002; 146(2): 21–28. PubMed Abstract \| Publisher Full Text\n\nArabmotlagh M, Sabljic R, Rittmeister M: Changes of the biochemical markers of bone turnover and periprosthetic bone remodeling after cemented hip arthroplasty. J Arthroplasty. 2006; 21(1): 129–34. PubMed Abstract \| Publisher Full Text\n\nAlbanese CV, Santori FS, Pavan L, et al.: Periprosthetic DXA after total hip arthroplasty with short vs. ultra-short custom-made femoral stems: 37 patients followed for 3 years. Acta Orthop. 2009; 80(3): 291–297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHua J, Walker PS: Closeness of fit of uncemented stems improves the strain distribution in the femur. J Orthop Res. 1995; 13(3): 339–46. PubMed Abstract\n\nBrodner W, Bitzan P, Lomoschitz F, et al.: Changes in bone mineral density in the proximal femur after cementless total hip arthroplasty. A five-year longitudinal study. J Bone Joint Surg Br. 2004; 86(1): 20–6. PubMed Abstract \| Publisher Full Text\n\nRahmy AI, Gosens T, Blake GM, et al.: Periprosthetic bone remodeling of two types of uncemented femoral implant with proximal hydroxyapatite coating: a 3-year follow-up study addressing the influence of prosthesis design and preoperative bone density on periprosthetic bone loss. Osteoporos Int. 2004; 15(4): 281–9. PubMed Abstract \| Publisher Full Text\n\nKorovessis P, Droutsas P, Piperos G, et al.: Course of bone mineral content changes around cementless Zweymüller total hip arthroplasty. A 4-year follow-up study. Arch Orthop Trauma Surg. 1997; 116(1–2): 60–5. PubMed Abstract \| Publisher Full Text\n\nSköldenberg OG, Bodén HS, Salemyr MO, et al.: Periprosthetic proximal bone loss after uncemented hip arthroplasty is related to stem size: DXA measurements in 138 patients followed for 2–7 years. Acta Orthop. 2006; 77(3): 386–92. PubMed Abstract \| Publisher Full Text\n\nKärrholm J, Anderberg C, Snorrason F, et al.: Evaluation of a femoral stem with reduced stiffness. A randomized study with use of radiostereometry and bone densitometry. J Bone Joint Surg Am. 2002; 84-A(9): 1651–8. PubMed Abstract\n\nPanisello JJ, Herrero L, Herrera A, et al.: Bone remodeling after total hip arthroplasty using an uncemented anatomic femoral stem: a three-year prospective study using bone densitometry. J Orthop Surg (Hong Kong). 2006; 14(1): 32–37. PubMed Abstract \| Publisher Full Text\n\nAlbanese CV, Rendine M, De alma F, et al.: Bone remodeling in THA: a comparative DXA scan study conventional implants and a new stemless femoral component. A preliminary report. Hip Int. 2006; 16(Suppl 3): 9–15. PubMed Abstract\n\nLerch M, Kurtz A, Windhagen H, et al.: The cementless Bicontact® stem in a prospective dual-energy X-ray absorptiometry study. Int Orthop. (SICOT) 2012; 36(11): 2211–2217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGruen TA, McNeice GM, Amstutz HC: \"Modes of failure\" of cemented stem-type femoral components: a radiographic analysis of loosening. Clin Orthop Relat Res. 1979; (141): 17–27. PubMed Abstract \| Publisher Full Text\n\nTrevisan C, Bigoni M, Cherubini R, et al.: Dual X-ray Absorptiometry for the evaluation of bone density from the proximal femur after total hip arthroplasty: analysis, protocols and reproducibility. Calcif Tissue Int. 1993; 53(3): 158–61. PubMed Abstract \| Publisher Full Text\n\nMirsky EC, Einhorn TA: Bone densitometry in orthopaedic practice. J Bone Joint Surg Am. 1998; 80(11): 1687–98. PubMed Abstract\n\nSpittlehouse AJ, Smith TW, Eastell R: Bone loss around 2 different types of hip prostheses. J Arthroplasty. 1998; 13(4): 422–7. PubMed Abstract \| Publisher Full Text\n\nVenesmaa PK, Kröger HP, Miettinen HJ, et al.: Monitoring of periprosthetic BMD after uncemented total hip arthroplasty with dual-energy X-ray absorptiometry--a 3-year follow-up study. J Bone Min Res. 2001; 16(6): 1056–61. PubMed Abstract \| Publisher Full Text\n\nNysted M, Benum P, Klaksvik J, et al.: Periprosthetic bone loss after insertion of an uncemented, customized femoral stem and an uncemented anatomical stem. A randomized DXA study with 5-year follow-up. Acta Orthop. 2011; 82(4): 410–416. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAldinger PR, Sabo D, Pritsch M, et al.: Pattern of periprosthetic bone remodeling around stable uncemented tapered hip stems: a prospective 84-month follow-up study and a median 156-month cross-sectional study with DXA. Calcif Tissue Int. 2003; 73(2): 115–21. PubMed Abstract \| Publisher Full Text\n\nBodén HS, Sköldenberg OG, Salemyr MO, et al.: Continuous bone loss around a tapered uncemented femoral stem. A long-term evaluation with DEXA. Acta Orthop. 2006; 77(6): 877–85. PubMed Abstract \| Publisher Full Text\n\nKobayashi S, Saito N, Horiuchi H, et al.: Poor bone quality or hip structure as risk factors affecting survival of total-hip arthroplasty. Lancet. 2000; 355(9214): 1499–504. PubMed Abstract \| Publisher Full Text\n\nGibbons CE, Davies AJ, Amis AA, et al.: Periprosthetic bone mineral density changes with femoral components of differing design philosophy. Int Orthop. 2001; 25(2): 89–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMunting E, Smitz P, Van Sante N, et al.: Effect of a stemless femoral implant for total hip arthroplasty on the bone mineral density of the proximal femur. A prospective longitudinal study. J Arthroplasty. 1997; 12(4): 373–9. PubMed Abstract \| Publisher Full Text\n\nSoininvaara TA, Miettinen HJ, Jurvelin JS, et al.: Periprosthetic femoral bone loss after total knee arthroplasty: 1-year follow-up study of 69 patients. Knee. 2004; 11(4): 297–302. PubMed Abstract \| Publisher Full Text\n\nHernandez-Vaquero M, Garcia-Sandoval MA, Fernandez-Carreira JM, et al.: Measurement of bone mineral density is possible with standard radiographs: a study involving total knee replacement. Acta Orthop. 2005; 76(6): 791–795. PubMed Abstract \| Publisher Full Text\n\nStilling M, Søballe K, Larsen K, et al.: Knee flexion influences periprosthetic BMD measurement in the tibia. Suggestions for a reproducible clinical scan protocol. Acta Orthop. 2010; 81(4): 463–470. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWindisch C, Windisch B, Kolb W, et al.: Osteodensitometry measurements of periprosthetic bone using dual energy X-ray absorptiometry following total knee arthroplasty. Arch Orthop Trauma Surg. 2012; 132(11): 1595–1601. PubMed Abstract \| Publisher Full Text\n\nSoininvaara TA, Harju KA, Miettinen HJ, et al.: Periprosthetic bone mineral density changes after unicondylar knee arthroplasty. Knee. 2013; 20(2): 120–127. PubMed Abstract \| Publisher Full Text\n\nData sheet of the device BMA TMMP D3A Medical Systems. Reference Source\n\nKim YH, Kim JS, Cho SH: Strain distribution in the proximal human femur. An in vitro comparison in the intact femur and after insertion of reference and experimental femoral stems. J Bone Joint Surg Br. 2001; 83(2): 295–301. PubMed Abstract \| Publisher Full Text\n\nWestphal FM, Bishop N, Püschel K, et al.: Biomechanics of a new short-stemmed uncemented hip prosthesis: an in-vitro study in human bone. Hip Int. 2006; 16(Suppl 3): S22–S30. PubMed Abstract\n\nWalker PS, Culligan SG, Hua J, et al.: The effect of a lateral flare feature on uncemented hip stems. Hip International. 1999; 9: 71–80.\n\nLeali A, Fetto J, Insler H, et al.: The effect of a lateral flare feature on implant stability. Int Orthop. 2002; 26(3): 166–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeali A, Fetto JF: Preservation of femoral bone mass after total hip replacements with a lateral flare stem. Int Orthop. 2004; 28(3): 151–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCossetto DJ, Goudar AI: Mid-term outcome of a modular, cementless, proximally hydroxyapatite-coated, anatomic femoral stem. J Orthop Surg (Hong Kong). 2012; 20(3): 322–6. PubMed Abstract\n\nD’Aubigne RM, Postel M: Functional results of hip arthroplasty with acrylic prosthesis. J Bone Joint Surg Am. 1954; 36-A(3): 451–75. PubMed Abstract \| Publisher Full Text\n\nAustralian Orthopaedic Association National Joint Replacement Registry Annual Report. Adelaide: AOA; 2010: 46–50.\n\nMinoda Y, Kobayashi A, Iwaki H, et al.: Comparison of bone mineral density between porous tantalum and cemented tibial total knee arthroplasty components. J Bone Joint Surg Am. 2010; 92(3): 700–6. PubMed Abstract \| Publisher Full Text\n\nZreiqat H, Crotti TN, Howlett CR, et al.: Prosthetic particles modify the expression of bone-related proteins by human osteoblastic cells in vitro. Biomaterials. 2003; 24(2): 337–346. PubMed Abstract\n\nWei X, Zhang X, Zuscik MJ, et al.: Fibroblasts express RANKL and support osteoclastogenesis in a COX-2-dependent manner after stimulation with titanium particles. J Bone Miner Res. 2005; 20(7): 1136–1148. PubMed Abstract \| Publisher Full Text\n\nRitchlin CT, Schwarz EM, O’Keefe RJ, et al.: RANK, RANKL and OPG in inflammatory arthritis and periprosthetic osteolysis. J Musculoskelet Neuronal Interact. 2004; 4(3): 276–284. PubMed Abstract\n\nKhosla S: Minireview: the OPG/RANKL/RANK system. Endocrinology. 2001; 142(12): 5050–5055. PubMed Abstract \| Publisher Full Text\n\nChilds LM, Paschalis EP, Xing L, et al.: In vivo RANK signaling blockade using the receptor activator of NF-kappaB:Fc effectively prevents and ameliorates wear debris-induced osteolysis via osteoclast depletion without inhibiting osteogenesis. J Bone Miner Res. 2002; 17(2): 192–199. PubMed Abstract \| Publisher Full Text\n\nUlrich-Vinther M, Carmody EE, Goater JJ, et al.: Recombinant adeno-associated virus-mediated osteoprotegerin gene therapy inhibits wear debris-induced osteolysis. J Bone Joint Surg Am. 2002; 84-A(8): 1405–1412. PubMed Abstract\n\nJosse R, Khan A, Ngui D, et al.: Denosumab, a new pharmacotherapy option for postmenopausal osteoporosis. Curr Med Res Opin. 2013; 29(3): 205–16. PubMed Abstract \| Publisher Full Text\n\nClohisy JC, Hirayama T, Frazier E, et al.: NF-κB signaing blockade abolishes implant particle-induced osteoclastogenesis. J Orthop Res. 2004; 22(1): 13–20. PubMed Abstract \| Publisher Full Text\n\nAbbas S, Abu-Amer Y: Dominant-negative IkappaB facilitates apoptosis of osteoclasts by tumor necrosis factor-alpha. J Biol Chem. 2003; 278(22): 20077–20082. PubMed Abstract \| Publisher Full Text\n\nClohisy J, Roy B, Biondo C, et al.: Direct inhibition of NF-kappa B blocks bone erosion associated with inflammatory arthritis. J Immunol. 2003; 171(10): 5547–5553. PubMed Abstract\n\nDai S, Hirayama T, Abbas S, et al.: The IkappaB kinase (IKK) inhibitor, NEMO-binding domain peptide, blocks osteoclastogenesis and bone erosion in inflammatory arthritis. J Biol Chem. 2004; 279(36): 37219–37222. PubMed Abstract \| Publisher Full Text\n\nAbu-Amer Y: Advances in osteoclast differentiation and function. Curr Drug Targets Immune Endocr Metabol Disord. 2005; 5(3): 347–355. PubMed Abstract\n\nSusva M, Missbach M, Green J: Src inhibitors: drugs for the treatment of osteoporosis, cancer or both? Trends Pharmacol Sci. 2000; 21(12): 489–495. PubMed Abstract \| Publisher Full Text\n\nViolette SM, Guan W, Bartlett C, et al.: Bone-targeted src SH2 inhibitors block src cellular activity and osteoclast-mediated resorption. Bone. 2001; 28(1): 54–64. PubMed Abstract \| Publisher Full Text\n\nGoldring SR, Gravallese EM: Bisphosphonates: environmental protection for the joint? Arthritis Rheum. 2004; 50(7): 2044–2047. PubMed Abstract \| Publisher Full Text\n\nSato M, Grasser W: Effects of bisphosphonates on isolated rat osteoclasts as examined by reflected light microscopy. J Bone Miner Res. 1990; 5(1): 31–40. PubMed Abstract \| Publisher Full Text\n\nCarano A, Teitelbaum SL, Konsek JD, et al.: Bisphosphonates directly inhibit the bone resorption activity of isolated avian osteoclasts in vitro. J Clin Invest. 1990; 85(2): 456–461. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPurdue PE, Koulouvaris P, Potter HG, et al.: The cellular and molecular biology of periprosthetic osteolysis. Clin Orthop Relat Res. 2007; 454: 251–261. PubMed Abstract \| Publisher Full Text\n\nSato M, Grasser W, Endo N, et al.: Bisphosphonate action. Alendronate localization in rat bone and effects on osteoclast ultrastructure. J Clin Invest. 1991; 88(6): 2095–2105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeedor JG, Quartuccio HA, Thompson DD: The bisphosphonate alendronate (MK-217) inhibits bone loss due to ovariectomy in rats. J Bone Miner Res. 1991; 6(4): 339–346. PubMed Abstract \| Publisher Full Text\n\nBalena R, Toolan BC, Shea M, et al.: The effects of 2-year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry, and bone strength in ovariectomized nonhuman primates. J Clin Invest. 1993; 92(6): 2577–2586. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMillett PJ, Allen MJ, Bostrom MP: Effects of alendronate on particle-induced osteolysis in a rat model. J Bone Joint Surg Am. 2002; 84-A(2): 236–249. PubMed Abstract\n\nLyons A: Effects of alendronate in total hip arthroplasty. Proceedings of the South African Orthopaedic Association. J Bone Joint Surg Br. 1999; 81(Suppl 3): 313.\n\nWilkinson JM, Eagleton AC, Stockley I, et al.: Effect of pamidronate on bone turnover and implant migration after total hip arthroplasty: a randomized trial. J Orthop Res. 2005; 23(1): 1–8. PubMed Abstract \| Publisher Full Text\n\nIm GI, Qureshi SA, Kenney J, et al.: Osteoblast proliferation and maturation by bisphosphonates. Biomaterials. 2004; 25(18): 4105–4115. PubMed Abstract \| Publisher Full Text\n\nIolascon G: Bisfosfonati e protesi articolari. Up-to-date 2005. In: Adami S (Ed.) Bisfosfonati Yearbook. 2005: 71–78. Reference Source\n\nRussell RG: Bisphosphonates: the first 40 years. Bone. 2011; 49(1): 2–19. PubMed Abstract \| Publisher Full Text\n\nCavalli L, Brandi ML: Targeted approaches in the treatment of osteoporosis: differential mechanism of action of denosumab and clinical utility. Ther Clin Risk Manag. 2012; 8: 253–266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOdvina CV, Zerwekh JE, Rao DS, et al.: Severely suppressed bone turnover: a potential complication of alendronate therapy. J Clin Endocrinol Metab. 2005; 90(3): 1294–1301. PubMed Abstract \| Publisher Full Text\n\nSomford MP, Draijer FW, Thomassen BJ, et al.: Bilateral fractures of the femur diaphysis in a patient in a patient with rheumatoid arthritis on long-term treatment with alendronate: clues to the mechanism of increased bone fragility. J Bone Miner Res. 2009; 24(10): 1736–1740. PubMed Abstract \| Publisher Full Text\n\nAspenberg P, Schilcher J, Fahlgren A: Histology of an undisplaced femoral fatigue fracture in association with bisphosphonate treatment: Frozen bone with remodelling at the crack. Acta Orthop. 2010; 81(4): 460–462. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTang SY, Allen MR, Phipps R, et al.: Changes in non-enzymatic glycation and its association with altered mechanical properties following 1-year treatment with risendronate or alendronate. Osteoporos Int. 2009; 20(6): 887–894. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDonnelly E, Meredith DS, Nguyen JT, et al.: Reduced cortical bone compositional heterogeneity with bisphosphonate treatment in postmenopausal women with intertrochanteric and subtrochanteric fractures. J Bone Miner Res. 2012; 27(3): 672–678. PubMed Abstract \| Publisher Full Text\n\nShane E, Burr D, Ebeling PR, et al.: Atypical subtrochanteric and diaphyseal femoral fractures: report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2010; 25(11): 2267–2294. PubMed Abstract \| Publisher Full Text\n\nWood J, Bonjean K, Ruetz S, et al.: Novel antiangiogenic effects of the bisphosphonate compound zoledronic acid. J Pharmacol Exp Ther. 2002; 302(3): 1055–1061. PubMed Abstract \| Publisher Full Text\n\nShane E, Burr D, Ebeling PR, et al.: Atypical subtrochanteric and diaphyseal femoral fractures: report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2010; 25(11): 2267–2294. PubMed Abstract \| Publisher Full Text\n\nAllen MR, Reinwald S, Burr DB: Alendronate reduces bone toughness of ribs without significantly increasing microdamage accumulation in dogs following 3 years of daily treatment. Calcif Tissue Int. 2008; 82(5): 354–360. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurr DB, Diab T, Koivunemi A, et al.: Effects of 1 to 3 years' treatment with alendronate on mechanical properties of the femoral shaft in a canine model: implications for subtrochanteric femoral fracture risk. J Orthop Res. 2009; 27(10): 1288–1292. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMashiba T, Hirano T, Turner CH, et al.: Suppressed bone turnover by bisphosphonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J Bone Miner Res. 2000; 15(4): 613–620. PubMed Abstract \| Publisher Full Text\n\nThompson RN, Phillips JR, McCauley SH, et al.: Atypical femoral fractures and bisphosphonate treatment: experience in two large United Kingdom teaching hospitals. J Bone Joint Surg Br. 2012; 94-B(3): 385–390. PubMed Abstract \| Publisher Full Text\n\nYoon RS, Hwang JS, Beebe KS: Long-term bisphosphonate usage and subtrochanteric insufficiency fractures: a cause for concern? J Bone Joint Surg Br. 2011; 93(10): 1289–1295. PubMed Abstract \| Publisher Full Text\n\nDas De S, Setiobudi T, Shen L, et al.: A rational approach to management of alendronate-related subtrochanteric fractures. J Bone Joint Surg Br. 2010; 92(5): 679–686. PubMed Abstract \| Publisher Full Text\n\nCross MB, Nam D, van der Meulen MC, et al.: A rare case of a bisphosphonate-induced peri-prosthetic femoral fracture. J Bone Joint Surg Br. 2012; 94(7): 994–7. PubMed Abstract \| Publisher Full Text\n\nBush LA, Chew FS: Subtrochanteric femoral insufficiency fracture following bisphosphonate therapy for osseous metastases. Radiol Case Rep. 2008; 3(4): 232. Publisher Full Text\n\nSayed-Noor AS, Sjödén GO: Subtrochanteric displaced insufficiency fracture after long-term alendronate therapy--a case report. Acta Orthop. 2008; 79(4): 565–567. PubMed Abstract \| Publisher Full Text\n\nKwek EB, Goh SK, Koh JS, et al.: An emerging pattern of subtrochanteric stress fractures: a long-term complication of alendronate therapy? Injury. 2008; 39(2): 224–231. PubMed Abstract \| Publisher Full Text\n\nGoh SK, Yang KY, Koh JS, et al.: Subtrochanteric insufficiency fractures in patients on alendronate therapy: a caution. J Bone Joint Surg Br. 2007; 89(3): 349–353. PubMed Abstract \| Publisher Full Text\n\nArnala IO: Salmon calcitonin (MIACALCIC NS 200 IU) in prevention of bone loss after hip replacement. Scand J Surg. 2012; 101(4): 249–254. PubMed Abstract\n\nAzria M, Copp DH, Zanelli JM: 25 years of salmon calcitonin; from synthesis to therapeutic use. Calcif Tissue Int. 1995; 57(6): 405–408. PubMed Abstract \| Publisher Full Text\n\nStock JL, Avioli LV, Baylink DJ, et al.: Calcitonin-salmon nasal spray reduces the incidence of new vertebral fractures in postmenopausal women. Three-year interim results of the PROOF study. J Bone Miner Res. 1997; 12: S149.\n\nAmmann P, Shen V, Robin B, et al.: Strontium ranelate improves bone resistance by increasing bone mass and improving architecture in intact female rats. J Bone Miner Res. 2004; 19(12): 2012–20. PubMed Abstract \| Publisher Full Text\n\nBuehler J, Chappuis P, Saffar JL, et al.: Strontium ranelate inhibits bone resorption while maintaining bone formation in alveolar bone in monkeys (Macaca fascicularis). Bone. 2001; 29(2): 176–9. PubMed Abstract \| Publisher Full Text\n\nDelannoy P, Bazot D, Marie PJ: Long-term treatment with strontium ranelate increases vertebral bone mass without deleterious effect in mice. Metabolism. 2002; 51(7): 906–11. PubMed Abstract \| Publisher Full Text\n\nHott M, Deloffre P, Tsouderos Y, et al.: S12911-2 reduces bone loss induced by short-term immobilization in rats. Bone. 2003; 33(1): 115–23. PubMed Abstract \| Publisher Full Text\n\nMarie PJ, Hott M, Modrowski D, et al.: An uncoupling agent containing strontium prevents bone loss by depressing bone resorption and maintaining bone formation in estrogen-deficient rats. J Bone Miner Res. 1993; 8(5): 607–15. PubMed Abstract \| Publisher Full Text\n\nBaron R, Tsouderos Y: In vitro effects of S12911-2 on osteoclast function and bone marrow macrophage differentiation. Eur J Pharmacol. 2002; 450(1): 11–7. PubMed Abstract \| Publisher Full Text\n\nCanalis E, Hott M, Deloffre P, et al.: The divalent strontium salt S12911 enhances bone cell replication and bone formation in vitro. Bone. 1996; 18: 517–23. PubMed Abstract\n\nTakahashi N, Sasaki T, Tsouderos Y, et al.: S 12911-2 inhibits osteoclastic bone resorption in vitro. J Bone Miner Res. 2003; 18(6): 1082–7. PubMed Abstract \| Publisher Full Text\n\nAmmann P, Badoud I, Barraud S, et al.: Strontium ranelate treatment improves trabecular and cortical intrinsic bone tissue quality, a determinant of bone strength. J Bone Miner Res. 2007; 22(9): 1419–25. PubMed Abstract \| Publisher Full Text\n\nLi Y, Feng G, Gao Y, et al.: Strontium ranelate treatment enhances hydroxyapatite-coated titanium screws fixation in osteoporotic rats. J Orthop Res. 2010; 28(5): 578–82. PubMed Abstract \| Publisher Full Text" }	[ { "id": "2668", "date": "20 Jan 2014", "name": "Jason K.F. Wong", "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 congratulate Loredana Cavalli and Maria Luisa Brandi on this really quite enjoyable review. For me, I found it quite comprehensive but there were some minor omissions such as the role of electrical charge on bone integration and role of tissue engineering and rapid prototyping on prosthetic design in the future. I would also have liked to have more information from the lessons learnt from failed implants in terms of the pathobiology at the interface. Otherwise it was a good review of the literature and worthy of publication.", "responses": [ { "c_id": "839", "date": "27 May 2014", "name": "Loredana Cavalli", "role": "Author Response", "response": "The authors would like to thank the referee for his response." } ] }, { "id": "4886", "date": "27 May 2014", "name": "Javad Parvizi", "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 a comprehensive evaluation of periprosthetic bone loss in mostly total hip arthroplasty that is helpful for orthopaedic surgeons. Currently, most orthopaedic surgeons only evaluate bone loss with plain radiographs. Having the knowledge to use DXA and BMA may be helpful in selective cases, although ordering advanced imaging on all patients may not be clinically necessary. Additionally, being able to combat bone loss with therapeutic treatments such as bisphosphonates and strontium ranelate may prove useful in a select patient population for a specified duration of time. Overall, this is a helpful article with cutting edge technology that may be useful for imaging techniques to better evaluate bone loss, and holds potential for evaluating future treatments to prevent and restore bone loss. Title/Abstract: Both are appropriate Imaging techniques for the evaluation of periprosthetic BMD loss:- This should include a section on x-ray imaging and CT imaging with metal subtraction.- What is the accessibility of BMA? What type of facilities have this? Are there studies comparing this technology with other imaging techniques with implants? Therapeutic strategies for enhancing bone mass recovery after arthroplasty:- Good sections on surgical and pharmacological considerations- Under “Pharmacological strategies to prevent aseptic loosening,” the authors mentioned the use of OPG and the soluble fusion protein RANK-Fc. However, there are no studies presented evaluating this in the setting of implants and only in the osteoporosis model. This may not be an appropriate to mention in this article (in comparison to the dominant negative IκB protein or the IKK inhibitory small peptide, which has been demonstrated to block particle-induced osteolysis).- It would be helpful to provide the mechanism of action of Strontium Ranelate. Conclusions: There should be a sentence on imaging, as mentioned in the article.", "responses": [ { "c_id": "840", "date": "27 May 2014", "name": "Loredana Cavalli", "role": "Author Response", "response": "The authors would like to thank the referee for his response.As regards BMA technique, it is not currently widespread for a clinical use, except in few research centers specialized in bone diseases, like ours at Orthopedic Trauma Center in Florence, Italy. As its entrance is very recent, and its potential applications are very wide, it is not yet available a protocol to compare this technique with other ones for peri-prosthetic bone study. So, at the end of the paragraph Imaging techniques for the evaluation etc., before the last sentence, we would add: \"Although not currently widespread, except in few research centers specialized in bone diseases, the employment of BMA...\" About metal subtraction option in X-ray and CT imaging, as it represents a real possibility to correct artifacts, we are preparing a brief paragraph to add: thank you for the suggestion! OPG and RANK-Fc were the first molecular strategy, so we would cite them, adding the sentence \"but the successful preclinical findings, i.e. the ability to prevent and reverse wear debris-induced osteolysis, have not been confirmed by clinical trials\". Regarding Strontium Ranelate, we would add: \"In particular, SrRan has been shown to enhance preosteoblastic cell replication and osteoblastic differentiation and to decrease abilities of osteoblasts to induce osteoclastogenesis, both through the calcium-sensing receptor and an increase in the OPG/RANKL ratio (new ref.: Brennan TC, Rybchyn MS, Green W, Atwa S, Conigrave AD, Mason RS. Osteoblasts play key roles in the mechanisms of action of strontium ranelate. Br J Pharmacol. 2009;157(7):1291–1300).\" In the conclusions, the sentence \"Dual-energy X-ray Absorptiometry allow to evaluate bone density around hip or knee prosthesis, and further data may be provided by the new device called Bone Microarchitecture Analysis.\" can be added after the second paragraph.We will provide a new version of the paper with all these corrections as soon as possible." } ] } ]	1	https://f1000research.com/articles/2-266
https://f1000research.com/articles/3-128/v1	17 Jun 14	{ "type": "Research Article", "title": "Expert interviews with science communicators: How perceptions of audience values influence science communication values and practices", "authors": [ "Paige Brown", "Rosanne Scholl", "Rosanne Scholl" ], "abstract": "Popular science communicators are a key link between scientists and publics, navigating the values of the scientific community and the perceived interests and values of readers. To do so, these professionals apply shared ideas about the role of science communication in society and about the factors that determine meaningful and newsworthy science stories. And yet we know little about the motivations and assumptions of audience values that underlie shared science communication values and criteria of story selection. Interviews with 14 popular science communicators writing in a variety of formats reveal that both their personal motivations and their perceptions of audience interests and values influence whether and how scientific research is translated into story.", "keywords": [ "“Science values detail", "precision", "the impersonal", "the technical", "the lasting", "facts", "numbers and being right. Journalism values brevity", "approximation", "the personal", "the colloquial", "the immediate", "stories", "words and being right now.” – Quentin Cooper of BBC Radio 4’s Material World1." ], "content": "\n\n“Science values detail, precision, the impersonal, the technical, the lasting, facts, numbers and being right. Journalism values brevity, approximation, the personal, the colloquial, the immediate, stories, words and being right now.” – Quentin Cooper of BBC Radio 4’s Material World1.\n\nWhile social science research has investigated the values of scientists in “hard” science fields toward production and communication of science, the values of journalists toward selection and production of news stories2 and human values on a fundamental level3,4, what appears to be lacking is a study of the interconnectedness of these value systems. Popular science communicators, or those including science journalists who translate scientific research into story for lay audiences, provide a case study in the intersection of these value systems. This is a professional group that balances adherence to the values of scientists, such as replication and hedging of scientific findings, while catering to the interests and values of lay audiences. Popular science communicators apply various norms and routines, especially those of science journalism, to science for the benefit of their readers. However, we know little about how popular science communicators perceive the interests and values of their audiences or how these perceptions affect how they cover science. How do they integrate what they perceive as audience interests and values with their own norms, routines and values as producers of popular science content? By answering these questions, we can develop a better understanding of what types and narratives of scientific research reach lay audiences through intermediary news stories, books, online science features, blog posts and other popular media formats. This is an investigation of what determines whether and how scientific research is translated into story. Specifically, we examine story selection and storytelling decisions of science communicators in the context of their subjective beliefs about readers.\n\nRelevance to readers is one news value that appears to be important in science communication. In a study of British journalists specializing in science, Anders Hansen found that this cohort “deploy conventional news-value criteria, but emphasize in particular the importance of a ‘relevance to the reader’ criterion in the selection of science news”5. A study of American science journalists published in 1979 revealed a trend in science journalism toward consumer-oriented coverage of science “…so that readers can answer the question ‘what does it mean to me?’”6. A focus on communicators’ values as well as audience interests in the production of news has received particular emphasis in science communication research5. However, this research has tended to ignore the influence of culture and “how specific aspects of science become newsworthy primarily because they relate to, or ‘resonate’ with, widely held cultural beliefs”5.\n\nThe main purpose of this study was to understand the experiences and perspectives of science communicators, both those in the field of journalism and those in complementary fields of science communication. When science communicators engage with their non-specialist audiences, they bring these perspectives to the interchange. We collected their personal accounts and explanations of practices, norms, routines and attributes of science storytelling during qualitative interviews. This study investigates specifically whether popular science communicators reflect fundamental and universal human values defined according to Schwartz’s3,4 value paradigm when selecting and telling stories about science. Additionally, this study investigates how perceptions of these values impact the narratives in which science communicators’ embed the science they cover.\n\nScience communicators’ use of traditional journalistic and storytelling norms and values in their construction of stories about science might be shaped by their perceptions of audience value orientations. News values, or the importance journalists attribute to various news factors which describe “why a topic is newsworthy and therefore published by the media”7, are already known to influence story selection8 and drive news coverage9. But we investigate the potential basis of these news values in perceptions of audience interests and values. For example, news values of unexpectedness, controversy and novelty might be associated with communicators’ subjective belief that their readers endorse motivational value types related to openness to change4 including hedonism, stimulation and self-direction. And story selection is not the only area of influence for science communicators’ perceptions of audience values. Perceptions of readers’ interests and fundamental value orientations may also color how communicators tell stories about science, which narratives they build around science and how they present scientific findings. The goals of this study were to identify the basic interests and value orientations attributed to lay audiences by science communicators, and to investigate how these perceptions might guide selection, production and content of science stories in a variety of media formats.\n\n\nLiterature review\n\nThe term “science communicator” as defined in this study includes non-fiction authors, news and magazine editors, journalists, university public relations (PR) writers and bloggers communicating primarily in areas of science including but not limited to basic physical sciences, biology, engineering, medicine and environmental science. Science communicators could be professional and/or freelance journalists writing about science full-time, editors of science magazines, university writers covering broad science research topics for lay audiences as well as news writers, non-fiction science authors and science bloggers in paid or non-paid positions. In some form or fashion, the science communicator translates science directly to lay audiences. It is important to note that in today’s media landscape, especially for science, storytellers often occupy diverse roles. For example, science PR practitioners may maintain science blogs that cater directly to lay readers, science bloggers often freelance for traditional news outlets, and science journalists often transition into roles as university science writers, bloggers, editors and book authors. In this landscape, science PR practitioners, bloggers and journalists all play roles in communicating scientific research, issues and events directly to readers online and offline as the traditional science journalism industry declines10. In fact, it is becoming increasingly difficult to differentiate the roles and work of science PR writers, science bloggers and science journalists10,11. It is for these reasons that we maintain that these communicators, although they occupy different spaces in the science news ecosystem and come from diverse backgrounds in science, journalism and mass communication training, all more or less actively consider the interests and values of lay audiences. These communicators also often share similar criteria of newsworthiness, despite their different backgrounds. For example, PR writers increasingly work closely with journalists and are often former reporters themselves. Science writers often float freely between the blogosphere and the worlds of online and traditional science journalism. For this study, interviewees were recruited broadly based on active practice in the communication of science to primarily lay audiences, through blogs, newspapers, books and other digital science news sites.\n\nWolfgang Donsbach describes12 four main factors that influence journalistic news decisions, or the decisions on what becomes story and how: (1) news values and specific news factors, (2) institutional objectives, (3) the influence of news sources including press releases and (4) the subjective beliefs of journalists. The two that are particularly relevant in this study include news values – or the importance attributed to various news factors – and the subjective beliefs of journalists about the values of their audiences. Although institutional objectives may help determine whether and how science communicators act upon their subjective beliefs and perceptions of audience interests and values, these beliefs and perceptions likely heavily influence the communication values and news factors that determine selection and production of stories about science.\n\nIn a 2013 issue of PNAS dedicated to the science of science communication, Thomas Dietz advocated bringing value research to the field: “Decisions always involve both facts and values, whereas most science communication focuses only on facts”13. Value theory research is based upon the idea that individuals orient their lives in society according to “a hierarchical arrangement of values, a rank-ordering of values along a continuum of importance”14. Building on Rokeach’s work on value typologies, Schwartz defined values as “the criteria people use to select and justify actions and to evaluate people (including the self) and events”3. According to Schwartz’s definition of universal human values adopted herein, a value is a “(1) belief (2) pertaining to desirable end states or modes of conduct that (3) transcends specific situations, (4) guides selection or evaluation of behavior, people, and events, and (5) is ordered by importance relative to other values to form a system of value priorities”4. Values are “desirable transsituational goals, varying in importance, that serve as guiding principles in the life of a person or other social entity”4. These goals have their basis in the biological needs of individuals, social interaction motives and requirements for group survival and function4,15. Schwartz developed a widely-used typology for the motivations behind values, which we will apply for the first time to science communicators.\n\nSchwartz found that universal human values fall into ten distinct motivational types. On a higher level, these motivational types nearly universally fall into two value orientation dimensions: self-interest (power, achievement and hedonism) versus altruism (universalism and benevolence), and openness to change (stimulation and self-direction) versus conservation/traditionalism (security, conformity and tradition)13. Starting with self-interest types, power values are derived from social interaction motives for recognition and dominance, include social power, authority and wealth. Closely related to power values, achievement values relate to personal success according to social standards, including intelligence, capability, ambition and influence over one’s peers or society in general.\n\nHedonism and stimulation values relate to physical needs and the desire for enjoyment in life, including happiness, comfort and pleasure (hedonism) and excitement, novelty, variation and challenge in life (stimulation). Also related to basic needs of the individual and social interaction and survival motives are security values, including diverse values such as health, social order, national and family security. These values refer to “safety, harmony, and stability of society, of relationships, and of self”4. In the openness to change category, self-direction values “refer to reliance on and gratification from one’s independent capacities for decision-making, creativity, and action,”15 and include curiosity, freedom, independence and the freedom to choose one’s own goals in life. On the other hand, tradition values refer to “respect, commitment, and acceptance of the customs and ideas that traditional culture or religion provide”4. Closely related to tradition values, conformity values refer to constraint according to social norms, and include self-discipline, obedience and honoring parents and elders.\n\nThe altruistic, or pro-social and group function, value types include universalism and benevolence. Universalism values, including social and environmental justice, unity with nature, inner harmony and broad-mindedness, refer to “understanding, appreciation, tolerance, and protection for the welfare of all people and for nature”4. Benevolence values, including love, meaning in life, honesty and helpfulness, refer to “preservation and enhancement of the welfare of people with whom one is in frequent personal contact”4. In some cultures, a related spiritual value type emerges, encompassing values including inner harmony and meaning in life.\n\nThis value typology will help in identifying the values science communicators perceive and cater to in their readers. This study investigates what motivational types and value orientations science communicators assume in their readers. These assumptions can have significant implications for narratives constructed around science in the media, and in turn public understanding of and attitudes toward science and its role in society13,16. It is important to note that the theoretical basis of this study includes fundamental human values as important aspects of science communication from the point of view of both content producers and consumers. Audiences will read and interpret stories about science based on preexisting value orientations whether or not science communicators acknowledge, understand or tap into these values.\n\nSchwartz’s universal human values typology has been applied in research on political attitudes and behavior4, environmental conservation17 and even journalists’ perceptions of their role in news work2. However this value typology has yet to be widely applied by mass communication and journalism researchers2. We propose that universal human values play a fundamental role in another value-related concept: news factors. Adoption of particular news factors, or the goals and decision factors used by journalists and editors to determine and/or ensure interest in and newsworthiness of a particular topic or story, may relate to perceptions of the interests, values and value orientations of intended audiences of that story.\n\n\nNews factors and newsworthiness\n\nSome communicators may supplement so-called human values with news values when they think about what their readers want. In journalism, all events are assumed to possess particular characteristics that make them more or less newsworthy, and it is these characteristics, gauged by journalists, that we refer to as news factors18. They make up the selection criteria used by journalists when presented with a number of story ideas overwhelmingly larger than the number they can actually pursue. According to Eilders, “[t]he more newsworthy an event is considered to be by journalists, the more likely it will be selected for publication, and the more likely it will be prominently presented”18.\n\nSome of the news factors that have been proposed to increase the likelihood of story selection include sensationalism19, frequency, proximity, importance to society, relevance to the audience, unexpectedness, oddity, drama, personalization/personification, negativism, controversy and facticity12,18,20,21. More recently proposed news factors include illustration of emotions, sexuality and visualization or the availability of graphical material18. But while some of these news factors may apply broadly across various areas of journalism, science journalists may endorse particular news factors over others. For example, surveys of British science journalists indicate that they look for potential science news that is “of ‘relevance to daily life’, ‘with a human angle’ […] or which is ‘weird and wacky’”5.\n\nIn his Psychology of news decisions: Factors behind journalists’ professional behavior, Donsbach went beyond the identification of specific new factors to a theory for why specific factors might be more powerful than others12. The author references Schulz in a partial answer to this question: “The only causal explanation for the existence of news values is journalists’ notions of audience interest”12. Other scholars echo this concept, that news factors are not only journalistic criteria, but also “general human selection criteria, deductible from the psychology of perception”18,22. In other words, anticipation of audience interests as well as audiences’ actual selection patterns underlie journalistic news values and specific news factors18.\n\nDonsbach enumerates two primary psychological explanations for broadly observed news decisions: journalists’ need for social validation of their perceptions, and journalists’ need to preserve pre-existing dispositions. The first explanation is a foundation for journalists’ heavy reliance on “shared” news values and standardized news factors, especially in cases where objective criteria are missing. The second explanation is a foundation for the impact of journalists’ perceptions of the interests and values of their audiences. While audience influence on news decisions can arise through institutional objectives, “journalists have their own conceptions of the audience and its taste”12. In this way, news factors and perceptions of audience interests and values may combine to influence journalists’ decisions of what to cover, and how.\n\nThe fact that science journalists have to be extremely selective in choosing amongst an almost infinite supply of story topics based on scientific research publications, press releases and other sources, potentially enhances the role of traditional news factors and perceptions about audiences on news decisions. For example, pre-existing perceptions of what audience members care about, or what their value orientations are likely to be, may guide story selection as well as storytelling.\n\nIn addition to news factors, nonfiction science writing often incorporates elements of storytelling including plot, strong characters, action, setting and language devices23. Many scholars have suggested that narratives are important means “to convey information in an accurate, attractive, imaginative and memorable way”24. Narratives, or stories, are easier to comprehend and more memorable than scientific facts24. Some of the narrative structures that have been proposed to enhance popular interest in science include stories of history, including historical accounts of research and scientist biographies25,26, stories highlighting a particular problem or drama, and stories about the struggles and successes of the scientists themselves. Stories should be exciting, elicit emotions, and should develop a connection between the reader and the subject24.\n\nBecause scientists have increasingly entered the public sphere in the communication of their work and science on a broader level27, and science journalists and bloggers are increasingly formally trained in the sciences10,28,29, we must consider the communication values of the scientific community as well as the journalistic and popular science writing communities in order to understand the storytelling decisions of science communicators. The most common theme in scientists’ motivations to communicate with the public is to educate, to improve understanding of science and/or to increase awareness30. However, other themes in scientists’ motivations to communicate with lay audiences include to inspire interest in science and to empower public decision-making30. Some of the ways that scientists seem to envision these goals being achieved include making science relevant to audience members, highlighting overarching concepts and helping audience members see how science and scientific data affect them personally30. These strategies seem to line up with some traditional journalistic news values, including importance to society, proximity and relevance to readers. However, scientists have also traditionally been skeptical about communicating science to broad public audiences, fearing that accuracy might fall prey to efforts at making science interesting, “whizz bang and glamorous”30. Scientists also tend to express the need for caution with controversial and preliminary research studies, particularly preliminary medical or drug research studies, when communicating findings to lay audiences31. It will be interesting to see how science communicators navigate these concerns while conforming to news values and efforts to engage lay audiences.\n\n\nMethods\n\nA single primary interviewer (PB), who is a Ph.D. student in the field of mass communication, conducted in-person, in-depth elite interviews with a total of 14 science communicators in North Carolina. The cities of Raleigh and Durham are often seen as hubs for science communication, hosting many prominent science bloggers, university science public relations officers (PIOs) and members of the society of Science Communicators of North Carolina, or SCONC. The final sample of communicators interviewed were recruited via personal e-mail invitation, starting with communicators the interviewer knew personally through social media (Twitter), followed by a snowball convenience sample of communicators suggested by initial interviewees. In e-mail invitations, interviewees were briefed on the goal of study “to learn from experts in the field how they craft their messages on science for the public” and invited to participate based on their expertise in the field of science communication. In responding to initial e-mail invitation responses, participants were also sent digital consent form files, which were signed, scanned and returned via e-mail, or signed and given to the interviewer in person.\n\nPotential participants suggested by “snowball” sampling were contacted strategically in order to get a mix of journalists, non-fiction science authors, bloggers and editors. One interviewee volunteered based on a Twitter announcement. Almost all science communicators who responded to a personal e-mail invitation ended up in the final sample. Out of 21 potential science communicators originally invited to interview via personal e-mail, 18 responded. One respondent no longer qualified as a science communicator due to a transition into foreign news reporting, and three declined to be interviewed due to scheduling problems. Representativeness of all science communicators was not a goal of the sampling procedure; this research is inductive and seeks to build theory rather than generalizable claims.\n\nInterviews took place in participants’ work offices, work-place conference rooms, or relatively quiet coffee shop locations convenient for participants. The semi-structured interviews, which typically lasted between 45–60 minutes, were digitally recorded and later transcribed in full by the single primary interviewer. Participants were assured of confidentiality in the use of the interview transcripts, and are thus identified here only by number (Interviewee 1–14 in chronological order). The interview transcripts and the notes taken by the interviewer during each interview form the basis of this analysis.\n\nIn-depth interviews were conducted according to methodology laid out by Lindlof and Taylor32, using a “guided introspection” interview protocol33. This type of interview taps into information and context that only the actor, in this case the science communicator, can provide, including perceptions, beliefs, values, and why he or she follows certain communication rules and routines. In-depth elite interviewing “stresses the informant’s definition of the situation, encourages the informant to structure the account of the situation, and allows the informant to reveal his or her notions of what is relevant33,34. Interview questions were kept broad and open-ended, with questions progressively but subtly prompting communicators to express their perceptions of audience interests and values (see Appendix for interview procedure).\n\nInterview analysis was approached predominantly using inductive and grounded theory methods: open coding for dominant emergent themes was followed by selective and deductive coding once key categories and themes – of audience engagement through story selection, content and presentation – had been identified. However, specifically for perceptions of audience values, interview analysis also involved deductive coding based on Schwartz’s universal human value typology3,4. Interview transcripts were coded and analyzed within ATLAS.ti (version 7.0.92), a software workbench for qualitative analysis of textual data. The science communication strategies and audience perceptions of the 14 science communicators were first examined as a group. Then the communicators were divided according to occupation (PR; freelance journalism & blogging; content management and editing) and educational background (formal science training vs. formal journalism training). In additional to these external trait divisions, analysis considered emergent divisions based on the communicator’s preferred science topic (“beat”), perceived role of science communication and other internal and work-based traits.\n\nScience communicators interviewed ranged in occupation from university public relations (4 interviewees; 1–10 years’ experience) or science PR consulting (1 interviewee; 19 years’ experience), to freelance science journalism in the form of newspaper articles and/or magazine articles and/or books (6 interviewees; 5–40 years’ experience), to science blogging (1 interviewee; 3 years’ experience), to content management and editing in the blogosphere (1 interviewee; 10 years’ experience) or in print (1 interviewee; 18 years’ experience). The gender breakdown of interviewees was 5/9 male/female, and interviewees ranged in age from 25 to 78 years old. One interviewee held an undergraduate degree in History of Science. One interviewee held an advanced law degree. Five interviewees held advanced degrees in science fields. All other interviewees held graduate and/or undergraduate degrees in journalism and other humanities fields.\n\n\nResults\n\n\n\nFollowing several initial rounds of open-coding and subsequent selective and deductive value-based coding of interview transcripts, codes of interest were categorized into four primary theme categories, according to important aspects of science communication brought up by interviewees: (1) news factors, or newsworthiness criteria, and institutional constraints; (2) messaging tactics and science narratives used to engage audiences; (3) perceived audience values (described in Table 1 and represented visually in Figure 1); (4) science communication values (described in Table 2). In the context of this analysis and the question of whether science communicators consider their audience’s values in selecting and telling stories about science, where applicable themes and codes are interpreted in light of how they relate to Schwartz’ value typology. Major and minor themes and codes described below are also quantified by the fraction of interviewees mentioning them (in parentheses).\n\nBolded specific values emerge as dominant themes in “audience” values perceived by science communicators.\n\n*Indicates a minor theme\n\nNews factors, or criteria of newsworthiness, that appeared to dominate science communicators’ selection and production of stories included (1) relevance or usefulness to the reader, characterized by applicability of scientific finding(s) to the reader’s daily life (mentioned by 12/14 interviewees), and (2) importance to society, characterized by answers to the questions “why is this [scientific finding] important?” and “why does it matter?” (9/14 interviewees). Related to these factors was the perceived need to show real-world applications, especially health, medical and/or clinical applications, of scientific research findings (8/14). One interviewee expressed this news factor as communicating to audiences that \"this could have implications for...\" Related to this was personalization, or referring to a particular person or personal story to show importance of a scientific finding for society (8/14). This might entail talking about a patient in a story about the pharmaceutical drug process.\n\nAnother dominant news factor was surprise or unexpectedness, characterized by astonishment, shock, or ideas that are counterintuitive, unexpected or conflicting with previous evidence (10/14). Two interviewees shared a special term for this – the ‘zag’:\n\n“Like ‘hun, that’s crazy, I never really would have guessed that that would be true,’ right? Um, so it’s those sorts of things. The… the ‘zag,’ it’s unexpected, it’s not really… it definitely counters what you thought would be, what’s going to come in the story. Yeah, so those are kind of the ones that do really well.” #5, Male, University PIO\n\nOf interest, news factors including relevance to the reader and applications of research were particularly emphasized by public relations communicators and freelance journalists covering biological science topics, health and medicine. However, these were either not mentioned or rarely mentioned by interviewees covering space, technology, engineering, physics, energy and environment and conservation issues. On the other hand, surprise factors were mentioned more frequently by interviewees covering the later topics (with the exception of environment and conservation). The area of science covered may alter the relevance of particular news factors to story selection and production, where physical or basic science research that can’t necessarily be made directly relevant to the reader most possess a high degree of novelty and unexpectedness to make a good story:\n\n“I mean, most people want to know, how is this going to directly impact my life, you know, and if you’re writing things for newspapers, or about health, or whatever, yeah, you can kind of see that a little bit more clearly. Um, but especially when it’s stuff like particle physics, or, you know, origins of the universe, or things like that, you know, it’s like yeah, there’s not really any immediate connection, but I think that people that read science are curious about that kind of stuff just for its own sake. So, um, in some ways just the fact that it’s something new is more of what you’re trying to convey, rather than immediate personal connection to, you know, subatomic particles (laughs).” #13, Female, Editor, Print\n\nOther important but less dominant news factors included scientific importance, or the importance of an event for scientific progress (8/12), proximity and localization (6/12), novelty or a certain threshold of news impact (4/12), the presence of graphical material to accompany a story (4/12), conflict or controversy (4/12), fairness to different views (4/14), economic relevance (4/12) and political or public policy impact (4/12). Minority expressed news factors included the presence of human interest or colorful people, and the presence of a news peg or a current event, mentioned specifically in the context of space history stories tied to spaceflight anniversaries, and environmental/conservation stories tied to current events. As might be expected, recognized traditional news values such as proximity and threshold were mentioned less frequently by science bloggers, blog managers and nonfiction book authors than by freelance journalists, university PR writers and editors.\n\nIn describing efforts to engage lay audiences in science stories, interviewees often mentioned the need to address the “why should I care?” question, to show readers why they should care about a particular scientific finding or event. According to one interviewee, the “second sentence of a press release has to address the question, ‘why does anyone care?’” (#2, Female, PIO). A majority of science communicators appear to address this question in part by showing how the science at hand relates to the reader’s daily life: “the idea of connecting it [the science] is to relate to something that people do every day,” (#4, Female, Science PR Writer). Especially with complex topics, relating a scientific finding or phenomenon to readers’ everyday experiences is often achieved by creating metaphors or analogies to things people already know (7/14) or talking about science in people’s own lexicons. For example, one interviewee gave the example of relating nanoparticle interactions to tying one’s shoestrings: “it just created this wonderful image that everybody knows. […] So it really is about, you know, just painting that very simple picture of something that, that people experience almost in their daily lives” (#10, Female, Freelance Journalist). Another interviewee who had been trained in science writing and taken a university course on science as literature expressly connected the use of analogies with storytelling devices that help audience understand and remember details35,36. The perceived importance of incorporating analogies to things people are already familiar with might have foundations in schema theory and the idea of conceptually-driven information-processing where, “[p]rior knowledge makes it easier to integrate new information and also raises the level of interest because the objects have already acquired meaning for a subject”18.\n\nStorytelling, or using elements of fiction including development of narrative, metaphors, characters, plot, tension and drama, emerged as a major theme in science communicators’ descriptions of their efforts to connect with lay audiences. All interviewed communicators mentioned at least one of these elements of fiction as important devices to engage audiences (with the exception of one research PR consulting agent trained in strategic communication):\n\n“Um, I think when you … recast science in words that … people would use at a dinner party, then they become engaged, and then they have the opportunity to … start thinking about science in their own lexicon, in their own … in their own bubble, if you will. And when they can do that, they do begin to care. And it may not be to a passionate level, you know, but it is at least to the point where, you know, they can, they can conceive of how science impacts their life.” #10, Female, Freelance Journalist.\n\nAnswering the “why should I care?” question for audiences was also associated with the institutional constraint, perceived especially by university PIOs and freelance journalists, of needing to address editors’ concerns and questions of “why do I care about this? Why does this matter?” (#5, Male, University PIO). Whether and how this question is addressed specifically, however, seems to depend both on the science topic being covered and the perceived value orientation of the reader.\n\nExplicit or implied perceptions of audience values were coded according to Schwartz’s value typology while allowing for emergent values via open coding. Our findings of interviewees’ perceptions of audience values in the selection and production of stories about science are shown visually in Figure 1, arranged according to Schwartz’s fundamental human value typology. In Figure 1, larger and bolded fonts represent values heavily reflected in science communicators’ work.\n\nSeveral dominant audience value themes emerged. Security values, including specifically individual health and family security and/or well-being, constituted an overwhelming majority of factors mentioned in response to question prompts “what are examples of stories people can relate to?” and “what do you think your audience cares about?” Several interviewees associated security values more generally with their readers’ tendency to have self-interested concerns and goals, confirming the idea that security values are related to basic needs of the individual3,4 and an egoistic value orientation37. Interviewees that connected security values with the need to engage readers’ self-interests largely worked in public relations or freelance journalism, in areas of medical and health reporting:\n\n“Well, health stories are obviously a winner, right? People care about themselves. […] again, all of the health topics. Cancer, heart disease, diabetes, obesity, I mean those all have immediate… immediacy for people, those are things they care about.” #5, Male, University PIO\n\nSeveral communicators also mentioned security values as guiding principles for their selection of potential stories:\n\n“[W]hen choosing a story to write for the lay public, it, you really do have to pick something that, if it’s a small development, or an esoteric development, it has to be something that impacts... a relatively widely-understood topic. Like if it’s, even if it’s just ah, ah, a gene modification, or a biomarker that’s, you know, rare or whatever, it has to affect something big, like Alzheimer’s, or breast cancer. You know, if it’s something small, it’s got to be part of a big problem, or people are going to not really care, because they are going to say, ‘well, what does this have to do with me?’” #10, Female, Freelance Journalist\n\nSecurity values of health and family security and/or well-being were also often mentioned in association with dominant news value themes including importance to society, relevance to the reader and showing (health and clinical) applications of research findings. However, even interviewees who didn’t stress traditional news values – one science blogging editor and one general science and energy nonfiction author who seemed to rely more on curiosity, entertainment and surprise to engage audiences – still mentioned security and health as basic audience values and concerns:\n\n“I had to actually read and learn some human and medical aspects of my field, beyond my, you know, [biological clocks] 101 lecture notes (laughs), you know, that I took 10 years before, where, you know, we spent maybe two weeks on humans (laughs). And, uh, the very basics. So I had to learn in order… because that’s what people were asking for.” #3, Male, Blog Editor\n\nA second dominant theme in perceived audience values included hedonism and stimulation values, two categories that often overlapped when mentioned during interviews. In the hedonism category, perceived audience values included happiness/enjoyment, comfortable life (“[t]hey are interested in anything that can make their life easier…” #1, Male, PIO) and entertainment (“when we read news […] I think we also want to be entertained” #10, Female, Freelance Journalist). In the stimulation category, perceived audience values included exploration and excitement (“I really work to convey excitement in and wonder at the topics” #11, Male, Journalist, Author). Hedonism values often overlapped with security and health values, especially in the case of one university PR writer who considered quality of life to be dependent on health and happiness (#1, Male, PIO). Communicators appeared to cater to both security and hedonistic values, such as comfort in life, when trying to engage wide audiences in science stories:\n\n“[I]f you’re trying to get that AP story that’s going to go in every local newspaper in the country, and then maybe hop across to Europe, etc. etc., then you want to explain why (emphasis) this particular scientific finding is relevant to you [the audience]. Is it gonna stop MRSA? Is it going to maybe lead to some revolutionary new treatments for cancer, is it going to help paralyzed people walk again? Um, is it going to make your food safer? Is it going to make, is it going revolutionize the way we think about clothing? Stuff like that. Hospital gowns. We redesigned a hospital gown, and we have gotten huge interest in that story. And you think, well nobody wants … the hospital is an uncomfortable, horrible place to be. The bottom line is that if you just be assured that your rear end is covered at all times (chuckles), that is one less thing to be miserable about. So it can be really simple stuff like that.” #2, Female, University PIO\n\nBelief that their readers value entertainment (in the hedonistic value category) and excitement (in the stimulation value category), on the other hand, was associated with the preferential selection of “cool” science topics. These topics included subjects such as “cockroaches, or robots, or an invisibility cloak,” (#1, Male, PIO) or science myths, science fiction fantasies or paleontology. These seemed to be topics that science communicators understood to not require extra effort to answer the “why should I care” question, to personalize the subject or to show relevance to daily life:\n\n“Then there are things that people are just interested in, like the cool science topics. Paleo. You don’t have to … ‘this T-Rex is relevant to you because…’ no, everybody just wants to imagine big giant toothy guys running around chomping the heads off of triceratops, that’s it. It’s ok (laughs). So, it depends on the subject. But if you are doing something that is a little bit more complicated, a new chemical formula… that contextualization helps – the personalization helps. Think about you as a person. Ok you’ve found this compound, why do I care? It’s a compound that ‘well oh if we coated the curtains and the surfaces in a hospital with it and then turned on the UV lights it nukes every bacteria. Hey! MRSA solved, right? It doesn’t get in the hospital anymore, because we’ve killed it. That kind of thing. So you know, people understand that, and they are like ‘oh, well that would be great!’ Take a little bit of worry off their mind about a hospital stay or whatever. It could be anything like that.” #2, Female, University PIO\n\nIt appears that the perceived audience values that science communicators address in their selection and production of stories depend in part on the subject or topic of science being covered. Science communicators believe that while “cool” science can feed into human values of excitement, more complicated science topics – molecular biology, genetics, engineering and technology – require tapping into self-oriented security or hedonism values in order to engage audiences. For example, one science blogger who covered space science history and space trivia rarely expressed perceptions of audience security values and did not mention a need to address relevance to reader or the “why should I care?” question. Instead, she relies on surprise and unexpectedness, reflecting human values of excitement and curiosity.\n\nCatering to entertainment and stimulation values was also a strategy used by some science communicators as a way of engaging audiences otherwise not interested in science. One freelance journalist and nonfiction science author in particular held a minority view that catering to audience self-interests by stressing relevance to the reader and explicitly answering the ‘why should I care?’ question involved a form of “talking down” to the audience (Interviewee #11). This interviewee instead works to incorporate a sense of amazement into his writing, assuming humans are inherently curious and interested in all things. This science writer emphasized the otherwise minority theme of universalism values among his readers, including broad-mindedness and a universal desire to seek an “understanding of ourselves and our universe” (#11, Male, Author). His interview introduced a unique science communication theme (Table 2) when he described that his job “is to be… the lighthouse, not the tugboat. Like, I illuminate so you can see what’s out there, rather than going and getting you and pulling you towards where you need to be, where I think you need to go” (#11, Male, Author). However, he also said that he realized that although he hopes that his audiences recognize that everything and everyone is connected, in global energy issues for example, readers are often motivated instead by egotistical concerns and issues closer to home, such as security and health. This is an interesting finding – while most science communicators assume their audiences tend toward the self-interested end of Schwartz’s human value spectrum, some work hard to activate self-transcendent values in their readers. To do this, interviewee #11 relies on entertainment and stimulation values as well as vivid storytelling:\n\n“I’m assuming everybody’s interested in everything, and they should be, and if they’re not, I’m so entertaining as a storyteller that I’m going to get them interested in it.” #11, Male, Author\n\n“Your job is to be the eyes and the ears of everybody else. Your job is to say, ha (excited) this is my job! I’m looking at nuclear waste in a pool, and I’m seeing it glow blue. And this guy’s explaining to me what Cherenkov radiation is, and that it’s beta particles travelling faster than the speed of light in water – in water, right, light doesn’t travel at the speed of light, it travels more slowly – beta particles travel slightly faster than light does in water, and so they, all kind of strange things happen, which leads to this blue glow, that’s the best I can do. But, um, like, and I’m writing it down, and I’m wearing a hard hat, and this is what I get paid to do! Oh my fucking god! And you have to be saying that, not every minute is like that, but generally, you know, that’s what your job is, and to convey that sense of amazement, and that understanding, to your audience.” #11, Male, Author\n\nSelf-direction values of curiosity, creativity and choosing one’s own goals in life were mentioned or implied by half of interviewees. Perceptions of these values seemed to be partly dependent on the science topic being covered (especially physical sciences, nature and environment topics, science history and science in culture), but also related to particular science narratives (including ‘untold stories’) and science communicators’ own values. For example, one interviewee, a science blog manager, stressed self-direction values in the context of his own view that science should always be communicated in its historical and culture context. When asked why history and culture were important aspects of science communication, he responded:\n\n“[F]irst it’s fascinating on its own. But I think it’s also useful knowledge, for a lot of people who are, ah, who have their own, ah, goals, like how to change one little aspect of culture for the better. So they can be activists for a cause, or something like that. […] So, you know, how technological and societal changes can affect how you change your movement, for betterment of some aspect of society. And, uh, and science and how it is done is often a reflection of that. Because every scientific finding, in some way, not just explains the world, but also affects the way we think about the world. So seeing how, you know, a very important paper from 50 years ago has changed everybody’s thinking about ‘X’ is one of the paradigms, like, you know, how can we do something like this? Maybe we should do science! Maybe we should research how people think, or how movements win causes, or something like that. So, there is definitely practical aspects, like know your history so that you don’t repeat it, also know your history to see what works and what doesn’t so that you can modify it for the current situation or to succeed.” #3, Male, Blog Editor\n\nAnother science communication value associated with perceived self-direction audience values, visible in the above quote, is the desire to inspire an interest in science, mentioned by half of interviewees (see Table 2).\n\nThose science communicators who emphasized that their readers tend toward self-direction values, including curiosity, also did not typically mention security or hedonism values. Interviewees who emphasized innate curiosity as a defining characteristic of their audiences – one space history blogger (#6), one general science freelance journalist and author (#11), and one a freelance journalist and science textbook author (#14) – generally did not emphasize news factors related to self-interest, including importance to society, relevance to the reader or health and clinical applications of research findings. However, perceptions of strong reader curiosity were associated with the science communication value of fixing misconceptions and educating (a major theme mentioned by 10/14 interviewees). This was true especially for one interviewee who worked as both a freelance journalist and a science textbook author. This interviewee emphasized targeting young readers’ curiosity for learning, assuming people relate to the “quest of discovery” or the “adventure” of scientific discovery (#14, Female, Freelance Journalist, Educator). Especially in this case, students and regular readers of science were expected to be innately curious and interested in \"science for science’s sake,” a striking departure from the dominant strategy of targeting self-interest values of security and health.\n\nA minor theme in perceived audience values emerged in the form of universalism and pro-social benevolence values. Universalism values mentioned by science communicators (7/14) included unity with nature, protecting the environment and environmental and social justice. When mentioned, these perceived audience values seemed to be strongly associated with another minor science communication value theme (Table 2): open-mindedness and broad-mindedness. This was especially the case for an open-coded audience value theme associated with universalism: Meaning of life. This emergent value involved the perception that people find it important to think of the broader, philosophical questions of life, including origins of life and the universe, and the meaning and purpose of our human lives. This value was mentioned by one interviewee in the context of a story about a physicist’s search for the Higgs Boson and universal human questions including “why are we here, where did we come from, what are the origins of, of everything we see around us” (#4, Female, Science PR Writer). Interviewee #11 expressed this value as a human desire to understand ourselves better:\n\n“And, so I think that’s the hidden part of what we’re doing, is by talking about what makes us tick, you’re trying to also show us, what makes us good, what makes us better. Because I think that’s our job, in all parts of our lives, is to make the world a better place than it was an hour ago, a lifetime ago, a month ago. And, this is how we do that. And ah, whether it’s by doing science so that we understand and can cure disease, or by making art so that we more understand what makes us tick as humans, or by telling stories in print or on radio or in video, so that we understand our human lives, and we can do a better job of listening to each other, and communicating to each other, and making the world a better place.” #11, Male, Author\n\nPerceptions of audience universalism values did not typically coincide with mentions of major news values. In some cases, these values were actually associated with discussion about the difficulties of engaging audiences with particular science topics including environmental and conservation stories. The perceived importance of catering to universalism values seems to be largely dependent on the personal philosophy of the science communicator and his or her own orientation toward this value category, which includes values of broad-mindedness, inner harmony, social and environmental justice (Schwartz, 1994).\n\nA final minor theme in perceived value orientations of audiences was endorsement of pro-social benevolence values. These perceived values, unlike universalism values, were associated with traditional news values. They were often mentioned in connection with news factors including importance to society and personalization, especially in the context of biological science, health and medical news. For example, several interviewees stressed telling the story of a patient in articles about a new cancer treatment or the drug discovery process: “You have to, you have to make it about humans, you have to make it about people, that’s what we care about” (#11, Male, Author). Benevolence values including responsibility and helpfulness were mentioned in the context of an emergent narrative of scientists as heroes (6/14):\n\n“I think a lot of people come to science with the assumption that people are trying to help. Scientists go into science because they want to find out why something works, so that they can maybe find out a way to make it work better, or use it to help people. I haven’t met many Dr. Evils.” #2, Female, PIO\n\nTradition and conformity categories of Schwartz’ (1992; 1994) value typology were nearly absent from science communicators perceptions of audience values. This absence was not surprising given the emphasis on stimulation and self-direction values. Tradition and conformity values constitute one pole of Schwartz’ (1994) conservation/traditionalism vs. openness to change value orientation, while stimulation and self-direction values constitute the other. This relationship is translated visually in Schwartz’s (1994) circular value structure diagram, adapted in Figure 1 for this study, where security, tradition and conformity value categories occupy a space at a 180° diagonal from stimulation and self-direction values.\n\nPower values (i.e. wealth) were implied only in the context of showing the economic relevance of scientific findings, especially when communicating to members of legislature on the outcomes of government funding of science. Achievement values were mentioned in the context of avoiding jargon or forms of condescension toward readers (writing in ways that would make the audience feel unintelligent or uninformed). Achievement values of success and ambition were also mentioned in the context of an emergent science narrative theme of humanizing science (8/14), specifically helping audiences to relate to scientists’ struggles to overcome obstacles in the name of science:\n\n“I think we relate to people more than we relate to machines. So if you put it in terms of, I mean, if you bring out a struggle that we can all relate to, like, trying to meet a deadline, I mean, every single person who’s ever gone to school has tried to meet a deadline, and, you know, known that there’s a lot at stake and consequences and all that bad stuff. You can take that same emotion of fighting to success and apply that to an engineer working on a rocket, and all of a sudden that story becomes something that like, yeah, I know what that feels like.” #6, Female, Blogger\n\nInterviewees occasionally mentioned value-related obstacles in their science communication work. For example, they perceived that readers find it difficult to connect certain fields of science such as mathematics (#2, Female, PIO), biochemistry and particle physics (#5, Male, PIO) to things they value. Science communicators also seem to believe that most readers don’t share the values of the scientific community, generally not understanding the need for replication and why it takes so long for a new pharmaceutical to go through the phases of clinical drug development (#7, Female, Freelance Journalist) or why scientists perform experiments on model organisms such as Drosophila and mice (#10, Female, Freelance Journalist). One common perceived obstacle involved tradition and spiritual values as agents of misunderstanding science. Five science communicators perceived respect for tradition and spiritual human values as antithetical to the understanding of scientific findings and the acceptance of scientific explanations for natural phenomena such as climate change and evolution.\n\n“there are people that think that evolution is not real, and that everything right now is exactly the way God made it, initially, and that is going to affect the way they view anything we write about paleontology, or bacteria, or human development. So, yeah, everybody has their own beliefs. The examples that I gave tend to be sort of more often faith-based. But those are going to limit your ability to reach those audiences. They are going to limit the ability of those audiences to place any new research findings in context; or rather the context in which they are going to place those research findings is going to affect their ability to understand and appreciate them.” #1, Male, PIO\n\n“these questions of, you know, who are we? Where did we come from? What are we doing? […] science is one way of answering those questions… religion is another way of answering those questions.” #4, Female, Science PR Writer\n\n“the biggest things we get responses on are things like when we write articles, you know, breaking things down like intelligent design, and then we get 300 letters, from people who like, who… people in Christian schools who make all their students write letters and things like that, and you are just like… (sigh)… you know… it reminds you the world is a very different place.” #13, Female, Editor, Print\n\n\nDiscussion\n\nThe results of this study extend previous findings and assumptions about news values in science communication by introducing concepts of how science communicators’ story selection and storytelling processes reflect perceived audience values. Twenty years after Hansen’s 1994 findings, science communicators still appear to often cater to ‘relevance to the reader’ criteria in selecting and producing stories. However, answering the ‘why should I care?’ question for readers is a tactic used primarily by communicators who assume that their audiences value security and personal well-being. These science communicators stress the need to show readers how science directly impacts daily life and health. However, they also acknowledge the dangers involved in doing so, including ‘hyping’ scientific results and fostering undue hope in new medical therapies, for example. Also, perceptions of a self-interested value orientation seemed more prevalent among science communicators targeting wide lay audiences and/or translating biological and life science research into story. Thus, assumptions about audience interests and values, and the need to cater to these assumptions, may vary with the individual communicator as well as with communicators who write news stories for wide audiences versus those who blog for what they perceive as niche audiences38. This should be investigated in future research.\n\nConversely, as we suspected at the beginning of this project, communicators who assume their readers endorse value types related to openness to change4 seem to rely more on news factors including surprise factors, unexpectedness and novelty, as well as innate human curiosity as a motivation for science news consumption. A relative minority of science communicators reflected benevolence and universalism value orientations among their readers in describing how they select and tell stories about science. However, it is interesting that those communicators who did also emphasized being personally motivated to inspire a broad interest in science ‘for science’s sake.’\n\nIt is particularly interesting that while self-direction values – including choosing one’s own goals and curiosity – occupy a motivational space distant from security values according to Schwartz’s openness to change vs. conservation/traditionalism value orientation scale (see Figure 1), science communicators relied on both these value categories simultaneously to engage readers. Communicators who select and produce stories about science based on the assumption that their readers endorse both of these value types equally might develop stories that conflict with the values of readers at either extreme of the openness to change vs. conservation/traditionalism orientation scale. However, at least in this study, it was often the case that for any particular story, communicators relied heavily on either perceptions of security values or perceptions of self-direction values, dependent on the scientific topic and their own communication motives, but not both in the same stories.\n\nMost of the science communicators interviewed for this study emphasized the use of narrative specifically, and elements of fiction in general, as important tools of the trade. Specific science narratives also seemed to reflect particular perceptions of audience value orientations. Personalization narratives, like showing a patient in a cancer treatment story, and narratives involving ‘scientists as heroes’ of disease, were associated with perceived security values as well as readers’ concern for other people they feel close to. Popular narratives highlighting the struggles and successes of scientists or the history of science24 were associated with perceptions that all people can relate to having self-chosen personal goals and overcoming obstacles to achieve them – a self-direction value orientation.\n\nThese results have significant implications for our understanding of science communication and what governs the translation of science research in the public sphere. Science communicators’ personal motivations may combine with their perceptions of readers’ fundamental values to determine what kinds of science topics are selected for storytelling and what story elements are emphasized. Some science communicators frame scientific findings as directly relevant to daily life, while others rely more on entertaining storytelling, and still others expect their audiences to meet them half-way based on an innate curiosity and motivation to explore the meaning of life. This choice may depend more on communicators’ own communication goals and perceptions of other people’s values than solely on their professionally shared journalistic news values.\n\n\nData availability\n\nfigshare: In-depth interview data on the values and practices of science communicators, doi: 10.6084/m9.figshare.105165739\n\n\nEthical considerations\n\nExemption approval for human subjects research was obtained prior to conducting qualitative interviews with science communicators from the Institutional Review Board at Louisiana State University. Questions about the ethical considerations of this research study can be addressed to Dr. Robert Mathews, Chair, 130 David Boyd Hall, Baton Rouge, LA 70803.\n\nIRB# E8312\n\nApproval Expiration Date: 5/6/2016\n\nDue to ethical considerations for the privacy of the science communicators interviewed for this study, the data file accompanying this manuscript does not include full interview transcripts. Select anonymized interview excerpts can be supplied upon request. Quantified mentions of all major and minor codes, per all 14 interviewees, are provided in associated spreadsheet (see data file).", "appendix": "Author contributions\n\n\n\nPB and RS conceived the study. PB designed the interview protocol and questions and carried out the research. PB prepared the first draft of the manuscript. Both 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\nThe author(s) declared that no grants were involved in supporting 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\nAppendix:\n\nI am going to ask you a series of questions in order to learn more from you as an expert in science communication, about current science communication practices as well as the best practices you have encountered. I am interested in your opinions, preferences, processes, and beliefs on this subject.\n\nCan you tell me a little bit about your background? How did you get into science communication?\n\nHow would you best describe your specific area of science communication? Do you specialize in communicating about a particular area of science?\n\nCan you describe the audience you write for, in your own words?\n\nCan you walk me through the process you go through when you write a science press release, story, article, etc? What goes through your mind as you write such a message for a lay audience, and especially what goes through your mind as you consider how to best engage with this audience?\n\nWhat do you think are some of the best ways to engage your audiences with your messages? What practices, topics, methods or message attributes do you use to best connect with your audiences? Conversely, what practices, methods or message attributes have you found don’t connect well with your audiences, and how have you changed or overcome these?\n\nWhat do you think your lay audiences care about, and how do you address these cares? Can you reflect on what pre-existing opinions and concerns you think your audiences might bring to your messages? Do you think you currently address these pre-existing opinions and concerns in your messages? If so, how do you do so?\n\nDo you consider the social and/or individual values of your non-specialist audiences in constructing your science communications? How does that influence your messages?\n\n\nReferences\n\nTurone F: The Number Needed to Inform-What we talk about when we talk of science journalism. Epidemiology, Biostatistics and Public Health. 2013; 10(1). Publisher Full Text\n\nPlaisance PL, Skewes EA: Personal and Professional Dimensions of News Work: Exploring the Link between Journalists’ Values and Roles. Journalism & Mass Communication Quarterly. 2003; 80(4): 833–848. Publisher Full Text\n\nSchwartz SH: Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. Adv Exp Soc Psychol. 1992; 25(1): 1–65. Reference Source\n\nSchwartz SH: Are There Universal Aspects in the Structure and Contents of Human Values? J Soc Issues. 1994; 50(4): 19–45. Publisher Full Text\n\nHansen A: Journalistic practices and science reporting in the British press. Public Underst Sci. 1994; 3(2): 111–134. Publisher Full Text\n\nDennis EE, McCartney J: Science Journalists on Metropolitan Dailies: Methods, Values and Perceptions of Their Work. J Environ Educ. 1979; 10(3): 9–15. Publisher Full Text\n\nBadenschier F, Wormer H: Issue Selection in Science Journalism: Towards a Special Theory of News Values for Science News? in The Sciences’ Media Connection–Public Communication and its Repercussions. Springer. 2012; 28: 59–85. Publisher Full Text\n\nMcQuail D: McQuail’s mass communication theory. Sage publications. 2010. Reference Source\n\nO’Neill D, Harcup T: News values and selectivity. The handbook of journalism studies. 2009; 161–174. Reference Source\n\nBrumfiel G: Science journalism: Supplanting the old media? Nature. 2009; 458(7236): 274–277. PubMed Abstract \| Publisher Full Text\n\nFahy D, Nisbet MC: The science journalist online: Shifting roles and emerging practices. Journalism. 2011; 12(7): 778–793. Publisher Full Text\n\nDonsbach W: Psychology of News Decisions: Factors behind Journalists’ Professional Behavior. Journalism. 2004; 5(2): 131–157. Publisher Full Text\n\nDietz T: Bringing values and deliberation to science communication. Proc Natl Acad Sci U S A. 2013; 110(Suppl 3): 14081–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRokeach M: The role of values in public opinion research. Public Opin Q. 1968; 32(4): 547–559. Publisher Full Text\n\nSchwartz SH, Bilsky W: Toward a universal psychological structure of human values. J Pers Soc Psychol. 1987; 53(3): 550–562. Publisher Full Text\n\nSchultz PW, Gouveia VV, Cameron LD, et al.: Values and their relationship to environmental concern and conservation behavior. J Cross Cult Psychol. 2005; 36(4): 457–475. Publisher Full Text\n\nSteg L, de Groot JIM: Environental Values, in The Oxford Handbook of Environmental and Conservation Psychology, S.D. Clayton, Editor. Oxford University Press: New York. 2012. Publisher Full Text\n\nEilders C: News factors and news decisions. Theoretical and methodological advances in Germany. Communications. 2006; 31(1): 5–24. Publisher Full Text\n\nLippmann W: Public opinion. Transaction Publishers. 1932. Reference Source\n\nSchulz W: Die Konstruktion von Realität in den Nachrichtenmedien: Analyse der aktuellen Berichterstattung (Freiburg and München: Alber). H. Viehrig/Mass Media and Catastrophe Prevention. How to Avoid the Crisis After the Crisis. 1976; 267. Reference Source\n\nSinger E, Endreny PM: Reporting on risk: How the mass media portray accidents, diseases, disasters, and other hazards. Russell Sage Foundation. 1993. Reference Source\n\nGaltung J, Ruge MH: The structure of foreign news the presentation of the Congo, Cuba and Cyprus Crises in four Norwegian newspapers. J Peace Res. 1965; 2(1): 64–90. Publisher Full Text\n\nLittmann M: Courses in science writing as literature. Public Underst Sci. 2005; 14(1): 103–112. Publisher Full Text\n\nNegrete A, Lartigue C: Learning from education to communicate science as a good story. Endeavour. 2004; 28(3): 120–124. PubMed Abstract \| Publisher Full Text\n\nSolomon J: Meta-scientific criticisms, curriculum innovation and the propagation of scientific culture. Journal of curriculum studies. 1999; 31(1): 1–15. Publisher Full Text\n\nGough N: Environmental education, narrative complexity and postmodern science/fiction. Int J Sci Educ. 1993; 15(5): 607–625. Publisher Full Text\n\nRead SJ, Miller LC: Stories are fundamental to meaning and memory: For social creatures, could it be otherwise. Knowledge and memory: The real story. Advances in social cognition. 1995; 8: 139–152. Reference Source\n\nBonetta L: Scientists enter the blogosphere. Cell. 2007; 129(3): 443–445. PubMed Abstract \| Publisher Full Text\n\nBubela T, Nisbet MC, Borchelt R, et al.: Science communication reconsidered. Nat Biotechnol. 2009; 27(6): 514–518. PubMed Abstract \| Publisher Full Text\n\nDavies SR: Constructing Communication: Talking to Scientists About Talking to the Public. Science Communication. 2008; 29(4): 413–434. Publisher Full Text\n\nTanona S, Schenck-Hamlin B, Lara A, et al.: Scientists’ Public Communication Values. Reference Source\n\nLindlof TR, Taylor BC: Qualitative communication research methods. Sage. 2010. Reference Source\n\nDrumwright ME, Murphy PE: How advertising practitioners view ethics: moral muteness, moral myopia, and moral imagination. J Advert. 2004; 33(2): 7–24. Publisher Full Text\n\nDexter LA: Elite and specialized interviewing. ECPR Press. 2006. Reference Source\n\nGreen MC, Brock TC: The role of transportation in the persuasiveness of public narratives. J Pers Soc Psychol. 2000; 79(5): 701–21. PubMed Abstract \| Publisher Full Text\n\nDahlstrom MF, Ho SS: Ethical Considerations of Using Narrative to Communicate Science. Science Communication. 2012; 34(5): 592–617. Publisher Full Text\n\nDe Groot JI, Steg L: Relationships between value orientations, self-determined motivational types and pro-environmental behavioural intentions. J Environ Psychol. 2010; 30(4): 368–378. Publisher Full Text\n\nLowrey W: Mapping the journalism–blogging relationship. Journalism. 2006; 7(4): 477–500. Publisher Full Text\n\nBrown PK, Scholl RM: In-depth interview data on the values and practices of science communicators. figshare. 2014. Data Source" }	[ { "id": "5153", "date": "31 Jul 2014", "name": "Sarah R. Davies", "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 an interesting topic: as the authors note, the way that communicators imagine their audiences will shape their output in significant ways. And I enjoyed what clearly has the potential to be a very rich data set. But I have some reservations about the adequacy of that data set, as it currently stands, given the claims the authors make; the relevance of the analytical framework(s) they draw upon; and the extent to which their analysis has offered significant new insights - by which I mean, I would be keen to see the authors push their discussion further. My suggestions are essentially that they extend the data set they are working with to ensure that their analysis is both rigorous and generalisable, and re-consider the analytical frame they use. I will make some more concrete comments below.With regard to the data: my feeling is that 14 interviews is a rather slim data set, and that this is heightened by the fact that they were all carried out in a single location, and recruited via snowball sampling and personal contacts. What efforts have the authors made to ensure that they are not speaking to a single, small, sub-community in the much wider category of science communicators? - a case study, if you like, of a particular group of science communicators in North Carolina? In addition, though the authors reference grounded theory as a method for analysis, I got little sense of the data reaching saturation. The reliance on one-off quotes, and on the stories and interests of particular individuals, left me unsure as to how representative interview extracts were. I would therefore recommend either that the data set is extended by carrying out more interviews, in a wider variety of locations (e.g. other sites in the US), or that it is redeveloped as a case study of a particular local professional community. (Which would open up some fascinating questions - how many of these people know each other? What spaces, online or offline, do they interact in, and do they share knowledge, for instance about their audiences? Are there certain touchstone events or publics they communally make reference to?)As a more minor point with regard to the data set and what the authors want it to do, there were some inconsistencies as to how the study was framed. On p.2 they variously describe the purpose as to \"understand the experiences and perspectives of science communicators\" and the goals as identifying \"the basic interests and value orientations attributed to lay audiences by science communicators\". Later, on p.5, they note that the \"research is inductive and seeks to build theory rather than generalizable claims\", while in the Discussion they talk again about having identified communicators' \"personal motivations\" (p.12). There are a number of questions left hanging: is the purpose to understand communicator experiences - in which case why focus on perceptions of audiences? Where is theory being built, and in what ways can this be mobilised in future work? The way that the study is framed and argued as a whole needs, I would suggest, to be clarified.Relatedly, my sense is that some of this confusion is derived from what I find a rather busy analytical framework. I was not convinced of the value of combining inductive and deductive coding: if the 'human value typology' the authors use is 'universal', then what is added by open coding? Or, alternatively, why let their open coding, and their findings from this, be constrained by an additional, rather rigid, framework? The addition of the considerable literature on news values to the mix makes the discussion more confusing again. I would suggest that the authors either make much more clear the value of combining these different approaches - building new theory outlining how they relate, and can be jointly mobilised in practice - or fix on one. (My preference would be to focus on the findings from the open coding - but that reflects my own disciplinary biases.)A more minor analytical point: the authors note that their interviewees come from slightly different professions, and communicate through different formats, have different levels of experience, and different educational backgrounds - but as far as I can see there is no comparative analysis based on this. Were there noticeable differences in the interview talk based on these categorisations? Or was the data set too small to identify any potential contrasts or themes? A note explaining this would be useful.My final point has reference to the potential that this data set has, particularly if it is extended and developed. I would like to encourage the authors to take their analysis further: at the moment, I was not particularly surprised by the ways in which the communicators referenced news values or imagined their audiences. But it seems to me that the analytical work is not yet complete. What does it mean that communicators imagine audience values and preferences in the way that they do - who is included and excluded by these imaginations? One experiment might be to consider what 'ideal type' publics are created in the communicators' talk. What are the characteristics of the audiences constructed in the interviews and - presumably - in the communicative products of interviewees? What would these people look like? There are also some tantalizing hints in the Discussion that are not really discussed in the Findings - of, for instance, the way in which communicator's personal motivations may combine with their perceptions of audiences to shape their products. How does this happen? These are, of course, suggestions. But my wider point is that the authors need to show more clearly what is original and useful in their findings - what it is, exactly, that will be important to other scholars in the field.I hope my comments make sense - please do not hesitate to contact me if not.", "responses": [ { "c_id": "925", "date": "31 Jul 2014", "name": "Paige Brown", "role": "Author Response", "response": "Dear Sarah R. Davies, reviewer: As one of the authors, I thank you sincerely for thoroughly reading and critiquing our study.I think a few clarifications need to be made in a revised version of this article, to address your comments. We need to clarify that the goal of this study was not to create a generalizable dataset or analysis of science communicators, but rather to begin an exploratory analysis into a theoretical area that has received relatively little research, i.e. how story selection and production, and by extension news values, may reflect more fundamental perceptions of audience values by science writers. The fact that this research is centered around understanding how story selection and production reflects the science communicator's perceptions of audience values is the reason that open coding was combined with deductive coding. By deductively coding for universal human values according to an established human value typology (in a multiple iterative fashion that we believe reached saturation), and seeing where these codes overlap with and how they compliment openly-coded categories of news values, storytelling devices, science communicator values and science communicator personal motivations, among other categories, we begin to build theory around how the production of stories about science for lay audiences reflects perceived fundamental human values. We don't seek to fully establish a theory of news values, story selection and production and fundamental human values, but rather to establish a new typology and concepts to be tested by future researchers. We are happy to provide additional interview abstracts to reveal the representativeness of the quotes we have chosen for coded categories.We might, in a revision, more clearly establish this as a case study of the experiences and perceptions of a particular group of science communicators in North Carolina. However, while many of these science communicators know \"of\" each other, through social media and some personal contacts, they were not trained in similar programs, do not have very similar communication backgrounds, and they are not communicating in the same areas of science, nor to the same audiences, nor in similar capacities. We don't believe there are any unifying factors between these science communicators that would lead them to select and produce stories about science in similar ways, other than the fact that they all operate within a U.S. centered media culture. In fact, because these communicators were chosen based on communicating in a variety of formats (books, digital news, blogging, science public relations), and on a variety of science fields or beats, they are likely to have very different conceptions of who their audiences are, what their audiences are interested in and what their audiences value. We think, for this reason, that any unifying characteristics in their perceptions of audience values are due to more fundamental processes in science communication within a U.S. centered media culture.In reference to your question of what is means that communicators imagine audience values and preferences in the way that they do, this must be explored in further discussion within this paper but also with further research by others. In particular, a clear dichotomy emerges from the data that shows that depending on the field of science covered (biology vs. astronomy, for example), communicators assume different value orientations among their audiences. However, the reflection of these value orientations has not previously been explored in any extensive way by sociologists of the news who've identified news values pertaining to science including 'relevance to the reader' and surprise or 'the wow factor.' In the future, sociologists of science news and science blogging could more fully explore why different news values and story production norms have applied to different science beats - perhaps because communicators covering these different fields of science assume (or must assume by the nature of their fields/beats) different basic value orientations among their readers. To our knowledge, human psychology value research has previously not been considered by sociologists in explaining how science news is framed, or why certain stories are selected over others.Thank you,-Paige Brown" } ] }, { "id": "5350", "date": "02 Sep 2014", "name": "Holger Wormer", "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 role of science communicators, their self-perception and their role in a growing “direct-to-consumer-culture“ of scientific institutions communicating their work to a broader audience is an increasingly relevant issue. Therefore, the approach of the authors to shed light on the opinions and values of science communicators has to be appreciated. The authors should be encouraged to continue in this direction. However, I would recommend strengthening the theoretical background and some of the related research questions, restructuring the analysis of the existing data set in a certain way, and thinking about extending the data set.Actually, the theoretical background of the paper is mainly based on Schwartz’ concept of universal structures of human values and the theory of news values. However, some theoretical approaches of science journalism and science communication (especially in the sense of science PR) must also be considered more intensively. This may be useful especially to avoid one weak point of the sample which is the rather generalized all-of-the-same definition of a “science communicator”. When the authors state that “it is becoming increasingly difficult to differentiate the roles and work of science PR writers, science bloggers and science journalists” this may be true from a non-expert audience point of view but clearly less from a normative perspective. As Rensberger (2009) has pointed out the era of just – as I would call it – “embedded science journalism” should be rather a point left behind in history than a suitable approach for today. Interestingly the authors themselves relativize their own initially painted picture of a more or less homogenous universal science communicator type in their own findings: They point out, for example, that “traditional news values…were mentioned less frequently by science bloggers, blog managers and nonfiction book authors than by freelance journalists, university PR writers and editors”. Or, in the final conclusion: “This choice may depend more on communicators’ own communication goals and perceptions of other people’s values than solely on their professionally shared journalistic news values.”It seems appropriate to focus more on these differences of the “communicators” which is subject of intense debates in other countries. For example, Martin Bauer Editor of “Public Understanding of Science” underlines that a [from institutional science PR] “independent science journalism is most needed” today (Bauer 2013). Also quite recently, all big German Academies of Sciences have published recommendations questioning the role and quality of science communicators of scientific institutions between authentic science communication and just marketing of a scientific institution (www.leopoldina.org/en/publications/detailview/?publication[publication]=580&cHash=fdc9fd538a91f7ccd59866ae6b3e5254 ).Considering such emerging debates and the important theoretical approaches concerning science communication and journalism (as a third frame additionally to the news value and human value theories) the authors may also think about rearranging and analyzing – at least – the three subgroups of their data set: Actually, 5 interviewees seem to be allocated to the public relation world of science, 6 to (freelance) science journalism. However, in the paper itself the exact role of the 2 interviewees from the blogosphere as a third possible subgroup (Are they blogging scientists or blogging journalist or blogging PR officers?) remains as unclear as the role of the interviewed print editor. If one or more of these subgroups seem to be too in-homogeneous and not consistent in itself it could be also necessary to extend the sample in a more systematically stratified way.As the summed up findings given in the paper for the sample as a whole are often not really surprising, a better carving out of possible differences among the proposed subgroups (journalists, science communicators from scientific institutions, blogger) may deliver even more interesting findings which are especially important considering the outlined theoretical background beforehand: What kind of science communication should and can easily be done by the science communication department of scientific institutions? Where (for example in cases of controversial issues in society) is a blogging scientist or an independent journalist preferable? Which values and anticipated news values are shared by each subgroup and which accordance is rather limited? And finally: How strong are they modulated by the treated scientific topic (e.g., health and life sciences versus astronomy versus engineering etc.), their anticipated audiences and other factors? (concerning the relationship between scientists & journalists & the public see also, e.g., Peters/Dunwoody et al.) – Of course, some of such analysis is already given in the existing version of the article but often in a rather hidden, implicit way. A more consistent multidimensional analysis alongside the expected and normative functions of the different subgroups may be an amazing approach for the further work in the field.As a minor point concerning the data set it could also be asked how reliable the reported self-perception of the interviewees is. Our own work has shown that there seems to be a certain discrepancy between what journalists believe that they consider as an important news value and what you can finally identify as news values in their publications. A proposal (to be included in this work or for a further development) would be to strengthen the reliability of the findings by combining the interviews with another methodological approach. One way to do so could be a qualitative content analysis including a small controlled sample of publications of each interviewee: Are the news values and human values that can be identified in their pieces of work the same that they have mentioned in the interviews? In any case it should become clear already in the title as well as in the abstract of the publication that we see a very interesting but highly explorative piece of work which has the main focus to develop new methods of looking on the role and self-perception of different (!) kinds of science communicators. Furthermore, it should become clear that the (preliminary) findings are still limited to a “case study from North Carolina” (or at least “a case study from the US”).Despite all my (as I hope not too critical) comments I am really looking forward to see a revised version of the paper which has the potential to fill a real gap in the field of science communication research.", "responses": [] } ]	1	https://f1000research.com/articles/3-128
https://f1000research.com/articles/3-101/v1	07 May 14	{ "type": "Opinion Article", "title": "Computational science: shifting the focus from tools to models", "authors": [ "Konrad Hinsen" ], "abstract": "Computational techniques have revolutionized many aspects of scientific research over the last few decades. Experimentalists use computation for data analysis, processing ever bigger data sets. Theoreticians compute predictions from ever more complex models. However, traditional articles do not permit the publication of big data sets or complex models. As a consequence, these crucial pieces of information have disappeared from the scientific record. Moreover, they have become prisoners of scientific software: many models exist only as software implementations, and the data are often stored in proprietary formats defined by the software. In this article, I argue that this emphasis on software tools over models and data is detrimental to science in the long term, and I propose a means by which this can be reversed.", "keywords": [ "Computers have become an essential tool in many aspects of science: they help with collecting and processing data from observations", "evaluating theoretical models", "and communicating with fellow scientists. In the course of the few decades in which scientists have used computers", "computing technology has changed very rapidly. These changes have permitted significant progress in many fields of science. However", "they have also lead to a shift of focus from scientific to technological issues", "as scientists eagerly applied new computing technology to study ever more complex systems. The most visible consequence is that today high performance often takes priority over reliable results in computational science", "even though few scientists would openly admit this preference." ], "content": "Introduction\n\nComputers have become an essential tool in many aspects of science: they help with collecting and processing data from observations, evaluating theoretical models, and communicating with fellow scientists. In the course of the few decades in which scientists have used computers, computing technology has changed very rapidly. These changes have permitted significant progress in many fields of science. However, they have also lead to a shift of focus from scientific to technological issues, as scientists eagerly applied new computing technology to study ever more complex systems. The most visible consequence is that today high performance often takes priority over reliable results in computational science, even though few scientists would openly admit this preference.\n\nRecently, some of the negative consequences of rushing forward at a fast pace have become too visible to be ignored1–3: mistakes due to insufficiently verified software, lack of reproducibility due to incomplete publication of data and codes and blind trust in software without a deeper understanding of the methods applied. The frenzy of becoming ever faster is slowly giving way to a more sober attitude that reinstates reliability and verifiability as the prime values of science. The Reproducible Research movement4 argues that reproducibility, one of the core principles of science, must be required in computational science as well as in other sciences, necessitating the publication of all software and data sets that are used in a computational study. From a somewhat different angle, the Open Science movement5, whose goal is access for everyone to the process of research, comes to the same conclusion. As a result of these efforts, publishing scientific data and software has become not only possible but straightforward, and journals are starting to encourage or even require such publication to accompany the traditional article that describes a study’s methods and results.\n\nIf the code and the input data of a computational study are published, anyone could repeat the computation and verify that it produces the published results. This is often referred to as replicability. At this time there is no agreement on whether replicability is a useful characteristic of a scientific study; the references 6 and 7 show two opposing points of view. Replicability limits fraud by proving that the authors can actually compute the results that they show in an article. It can also be seen as a proof of quality assurance, because it demonstrates that the authors have recorded their complete computational workflow, which is not yet common practice. However, replicability does not mean that the authors did what they describe in their article, nor does it help the readers to develop a better understanding of the methods that were applied. Minimal replicability (such as making available a virtual machine image that runs the computation) doesn’t imply openness either, as readers cannot apply the published methods to different situations, or analyze the data using their own methods.\n\nReplicability is clearly not the same as the traditional notion of reproducibility in science. The latter requires that other scientists design their own experiments or computations, which incorporate the key elements of the original work but differ in points considered unimportant, and obtain similar results. In contrast to replication, reproduction of a scientific study adds new information that helps to identify what matters and what doesn’t matter for obtaining a specific result. Making a computational study reproducible thus requires explaining the methods behind it in a way that clearly states which aspects are considered important.\n\nThe central question in computational science is: why should we trust the results of a non-trivial computation? We all know from experience that software has bugs, and we also know that the use of computers is subject to frequent human mistakes. Moreover, computational scientists should be aware of the complexity of their software, and thus should be concerned that it might not do what they believe it does. Creating trust in computational results requires validation at all possible levels: we need replicability and reproducibility, and also an increased effort to explain our computational models and methods to our peers. The fundamental problem is that scientific software is much too complicated to be an efficient way to communicate these models and methods, and no other precise representation is available. A detailed understanding of what a given piece of software does is often limited to the software’s authors.\n\nIn this article I will explain why the current situation is unsatisfactory, and propose approaches for improving it. I will illustrate my explanations with examples from my own field of research, which is biomolecular simulation. However, after many discussions with computational scientists from other application domains, I conclude that the situation is very similar wherever computers are used for tasks that are impossible to do manually.\n\nIn order to make my point clear, I will first give a summary of the role of models in science, and of the role of computation in scientific models. This will set the stage for the following discussion on the current state of scientific software. I will then propose concrete actions that can be taken to improve the situation and outline the benefits that we can expect from them.\n\n\nModels\n\nThe central concept presented in this article is the notion of the scientific model. The role of models in science has been the subject of much debate among philosophers of science8. A good general overview written by a scientist for scientists9 and an illustration in the context of physics education10 have been given by Hestenes. A scheme of the process of scientific research (see Figure 1) illustrates the fundamental role of scientific models: science can be summarized as a process whose inputs are the data obtained from scientific observations, and whose output is a set of models with the associated validation information. As more data become available, new model/validation pairs are produced, which may be refinements of older models, but also completely new models. The defining characteristic of a scientific model is that it can be used to deduce verifiable statements about observable aspects of nature, which makes it possible to test and refine a model using new data from subsequent observations.\n\nThe main input is data from observations, the main output are models with associated domains of validation.\n\nModels do not necessarily need to be quantitative. The metabolic pathways in biochemistry are a well-known example for non-quantitative models. However, in the context of computational science, nearly all models are quantitative, as they predict numbers that are compared to the numbers obtained from the actual measurements. In the following, I will limit the discussion to quantitative models.\n\nThe models most frequently discussed in the context of scientific research are those for the systems in nature that we try to understand. However, we also use physical models to describe the instruments we use to make observations, and non-physical phenomenological models to account for the aspects that we do not understand in detail. The most common models in the last category are the statistical error models, such as the very frequently (and usually silently) made assumption that an observed value is the “real” value plus an “experimental error” described by a Gaussian probability distribution. Computational studies exploring models for systems in nature are called “simulations” and are often performed on models believed to be accurate, with the goal of obtaining information that is difficult or impossible to obtain from the observation. Simulations that include a model for the scientific instruments are often labeled as “virtual experiments”. Computational studies applying statistical models to the data are called “data analysis” and typically have the goal of determining a set of model parameters that best describe the data resulted from the observation or simulation. The arguments I present in this paper apply to all of these categories.\n\nMany scientific models are formulated in the framework of a theory which defines the general rules for a large class of models. An example is classical mechanics, which is a theory describing the dynamics of systems of point masses or finite-volume rigid bodies. Within the framework of classical mechanics, a model for a concrete system can be defined by a single function called a Hamiltonian. Theories play an important role in the most mature fields of science (e.g. physics) but are not essential for defining models. Younger disciplines, e.g. systems biology, construct models in a more ad hoc fashion without a clear underlying theory. Yet another approach is the construction of models derived from several theories in a multidisciplinary setting, e.g. in climate research. For the aspects that I discuss in this article, it does not matter if a model is developed in the context of some theory.\n\nComputable models are the models that are of prime interest in computational science. A computable model is a computable function, as defined in computability theory11, whose result can be compared to data from observations. Since validation requires the comparison of concrete results with observed data, one would expect that all quantitative models in science are computable models. Surprisingly, this is not the case. In fact, most mathematical models used in science are not computable.\n\nConsider, for example, the description of the solar system in terms of classical mechanics that goes back to Isaac Newton: a set of point masses (the sun and the planets) interacting through Newton’s law of gravitation and moving according to Newton’s laws of motion. The latter are differential equations for the positions and velocities of the celestial bodies. Together with a set of parameters obtained from observation (for example, the positions and velocities of all celestial bodies at a given moment in time), these equations determine the positions and velocities at any time in the past or the future. However, they do not provide a recipe for computing the actual numbers that could be compared to observations. An additional approximation is needed to obtain a computable model. For the simplest case of a system of only two celestial bodies, an analytical solution of the differential equations can be obtained. But this solution contains transcendental functions (sines and cosines), which are not computable and therefore must be replaced by computable approximations, e.g. power-series expansions. When three or more celestial bodies are included in the model, no analytical solution is available and the differential equations must be approximated by finite difference equations12. The development of computable approximations to the Newtonian model of celestial dynamics remains an active topic of research (see e.g.13). More generally, one can consider the whole field of numerical analysis as dedicated to constructing computable approximations to non-computable mathematical models.\n\nIt may seem surprising that most mathematical models used in the most mature domains of science do not strictly speaking deserve the label “scientific”, because they cannot make predictions that are immediately comparable to observed data. The explanation is that computation was for a long time considered a menial task not worthy of the attention of a distinguished mathematician or scientist, who should concentrate on mathematical and logical reasoning. As Dowek14 explains in a fascinating account of the interplay of reasoning and computation in mathematics and logic in the course of history, the important role of computation in formal reasoning has become clear only during the 20th century. While today it is generally accepted by mathematicians and logicians, most theoreticians in the natural sciences still consider computation an inferior approach to exploring scientific models, which is only used out of necessity when other techniques have failed. I suspect that this lack of interest by “real theoreticians” for the computational aspects of science might have contributed to the problems that I have outlined in the introduction. This would also explain why computational training is still largely absent from the science curricula around the world.\n\nScientific models can be written down in many ways: mathematical equations, diagrams, plain language, etc. The same model can be represented by different notations. For example, in principle any mathematical equation could be replaced by a verbal description. Computable models can be expressed in any Turing-complete formal language, and in particular in any of the commonly used programming languages, making them the most precise and unambiguous scientific models. The very fact that a program runs and produces results proves that the model specification is complete and unambiguous, assuming that the computing system itself (hardware, operating system, compiler, etc.) works correctly. The utility of computation in the process of understanding and documenting science has been pointed out by Sussman and Wisdom15, but is not yet widely recognized in the scientific community. A nice illustration from an engineering domain (the design of musical instruments) is given by Mairson16, who designed a computable notation for describing the geometrical constructions that have been used for a few centuries to construct string instruments. His notation is meant to be both a set of instructions for a computer and a precise and unambiguous description for human readers.\n\nA final important point about computable models is the importance of correctly identifying, understanding and documenting approximations. Scientists frequently make approximations to computational models without recognizing them as such, and therefore do not document these approximations in their publications. A good example is the use of finite-precision floating-point numbers in place of real numbers. Most scientists would consider this a technical necessity in implementing a model on a computer, and therefore an implementation detail of computational software. However, floating-point numbers have properties that differ significantly from real numbers (for example, addition and multiplication are non-associative), and the finite precision necessarily changes the results of the computations. In general, any modification to a computer program that changes its results implies an approximation to the original computational model. This also includes techniques such as lossy compression of output data, which again are usually considered implementation details.\n\nIn summary, computational science involves working with computable scientific models, which are either constructed from first principles or more frequently as approximations to non-computable models. A publication describing a computational study should contain a full description of the models that were actually used in the computations. For the models derived as approximations, this means that the final approximation, though preceding steps in the derivation, should also be given in order to document the process. Computable models can be expressed unambiguously in a Turing-complete formal language. A suitable Turing-complete language should be the preferred form for publishing models.\n\n\nTools and methods\n\nScientists use a variety of tools to gather observational data, explore the predictions of models and perform comparisons between them. I use the term “tool” in a general sense that includes both physical objects (e.g. microscopes, lasers, etc.) and mathematical theorems or procedures (e.g. calculus or algebra). Both computers and the software that runs on them are thus considered tools. Tools are evaluated by how well they help us in getting a job done, which leads to criteria such as precision, performance, efficiency, convenience and price. In scientific publications, the tools are described in the “Methods” section. A computational method corresponds to running one or more software tools with specific input parameters.\n\nPeople using tools, not only in science, develop a mental model of how the tools work and what they do. Such mental models are mostly empirical and are developed by training and experience. They are personal and not formalized in any way. There is no fundamental difference in how we form mental models of a car, a microscope, and a text editor running on a computer. Our mental models are limited to the aspects of the tools that we have to know, and they do not include the tools’ inner workings or construction details. For example, to drive a car, we need to understand accelerating, braking and steering, but not the process of combustion in the engine. Similarly, we can use a microscope or a text editor with far less knowledge than it takes to design and build one. However, the domain of application and the precision that we can expect from the results are part of the mental models that scientists need to have for their tools.\n\nWhile tools are indispensable for conducting science, they are not considered as part of the outputs of science, which consist of validated models. Articles documenting scientific studies describe the tools and methods that were used in the experiments or computations in order to permit readers to judge the pertinence of the conclusions drawn from the outputs. The development of new tools is also described in scientific publications because these tools are important products of the scientific research process. Nevertheless, these two aspects (tools and outputs) should be kept separate. The conclusion of a scientific study needs to be independent of a specific tool to deserve the name “scientific”. Another scientist should be able to reach the same conclusion using different tools, which is part of the requirement of reproducibility.\n\nIn computational science, the distinction between models and methods is not always very clear, because both take the form of algorithms. Some disciplines, e.g. bioinformatics, are very methods-oriented and rarely refer to models. A bioinformatician is more likely to propose a “method to predict protein folding” than a “model for protein folding”. This is partly due to differences in scientific jargon among disciplines, but it also reflects deeper issues concerning the role of computing in science. The global minimum of a knowledge-based potential for proteins is clearly a scientific model for a native structure. It is even a computable model in the sense of computability theory, in that there are known algorithms that can find the global minimum in finite time to any specified precision. However, that finite time is so long on today’s computers that the global minimum cannot be computed in practice. Bioinformaticians therefore construct heuristic methods that find structures close to the global minimum rapidly in the majority of cases. If these heuristic methods are deterministic, they should be considered approximations to the original model. This is not an option for heuristic methods that involve random choices, because they do not produce a unique result for a given input and therefore do not qualify as scientific models. The role of such probabilistic heuristics in computational science remains to be clarified.\n\n\nThe double role of scientific software\n\nThe dominant role of software in our lives is the role of tools. A computer program does something: play videos, manage bank accounts, simulate protein dynamics, etc. A software is developed explicitly for doing something, and is evaluated by how well it performs the task. In most situations where software is used, there is a clear distinction between the software as a tool and the content that the tool works on. A video player is distinct from the movies it plays, and this distinction is visible to everyone: there is one file on the computer for each movie, and one file (or set of files) for the video player. The same video player can play many movies, and for each movie file there are multiple computer programs that can play them. The same clear distinction holds between the software that manages bank accounts and the databases that contain the actual data. However, it does not hold for the simulation of protein dynamics. A simulation is the computation of a prediction from a model, but there is no computer file that holds a model for protein dynamics, and another file that holds a simulation program. The model is an integral part of the simulation program. The files read by that program contain some of the parameters of the model (e.g. the initial structure of the protein), but not the model itself. There is no clear separation between the tool and the model it operates on.\n\nThis fusion between models and tools in computational science is problematic because models and tools have very different roles in science and are evaluated according to very different criteria. Blurring the distinction leads to a number of undesirable consequences:\n\nLack of understanding: In the theoretical sciences, researchers should know and understand in detail the models they apply. These models are shared by a research community, and formalized using a suitable standard notation to reduce ambiguity in communication. Scientists do not have the same detailed understanding of their tools. Researchers using scientific software (as opposed to those who develop it) work with an empirical mental model of that software, as explained above. When scientific models are hidden inside the software, the higher level of understanding required for them becomes very difficult to develop. As a consequence, researchers cannot make an informed decision between different models and often choose the more convenient or more efficient program, regardless of the model that it implements.\n\nLack of verification: New tools should be tested by running them on well-known models as test cases, for which they should produce exactly the same results. New models should be tested by comparing them with well-known ones, using exactly the same tools. Software that inextricably combines complex models and complex technology becomes nearly impossible to evaluate. Moreover, formal proofs can be used to validate software tools against a formal specification. But formal proofs cannot handle scientific models (they are validated against observations), and therefore they cannot handle tools with built-in models either. Tools can only be validated using formal proofs if they work on models that are external.\n\nInterdependence: Models and tools should evolve independently: models are improved with the progress of science, whereas computational tools are improved following changes in computing technology, or simply by investing more efforts. When the models are part of the tools, it becomes difficult to distinguish an improved tool from an improved model. Moreover, changes to the tools for technical reasons (i.e. accelerating a computation using Graphics Processing Units (GPUs)) often require approximations to the embedded models, which tend to remain undocumented because they are not recognized as such.\n\nAll these consequences can be observed in the field of protein simulations. It is generally accepted in the protein simulation community that it is impossible to obtain the same numbers for a given system from two different simulation programs (lack of verification). Most scientists understand that this is due to no two programs implementing exactly the same model. However, few if any practitioners are able to explain how exactly these models differ (lack of understanding). It is also considered inevitable that different versions of the same program, or even two executables compiled with different compilers or compiler options, produce slightly different results for what should be the same model (interdependence).\n\nIt is important to understand the practical differences between a computable scientific model and a software tool. From the point of view of theoretical computer science, both are programs and both are expressed in a Turing-complete language. However, the model specifies just the result of a computation. The software tool defines how to perform a computation efficiently on data read from and written to permanent storage, within the constraints of a given physical computer. This requires handling aspects such as the use of resources (memory, CPUs), I/O, and possibly parallelization. A practically useful software tool also requires attention to the user interface, to file formats, and other tool-specific characteristics. In a typical scientific software tool that integrates models, the vast majority of the source codes is dedicated to these technical aspects, to the point that it can be difficult to identify the models in the source code.\n\nIt is interesting to analyze why the fusion of models and software tools is possible and how it occurred. In the cases of video players and bank account management cited above, the separation of tools and data seems evident. The tools consist of instructions for the computer, the data is ultimately just a sequence of numbers. Anyone who has written simple programs is able to see at a glance the difference between software (text files containing instructions in a programming language) and data (tables of numbers and text for the bank accounts, compressed binary files for movies). The archetype of a scientific model is a set of mathematical equations. This seems much more similar to a program than to data, all the more since most programming languages provide syntax for mathematical formulae that look similar to written maths. Moreover, as explained above, computable models actually require a Turing-complete notation. A programming language is thus a natural fit: it is very straightforward to translate a computational model into a program code. On the other hand, it is not at all straightforward to write a program that reads in a scientific model as it would read in “normal” data. So it seems that scientific models are in fact programs rather than data.\n\nHowever, the distinction between “program” and “data” doesn’t stand up to scrutiny. Programs are data. They are stored in files, can be copied around, e-mailed, etc., just like any other piece of data. Compilers read source code files as data and transform them into executables, which is just a conversion of data into another form. The distinction between programs and data that seems so obvious to computational science practitioners is just a historical accident. The programming language Fortran17, which made large-scale scientific computing possible in the late 1950s, made this distinction for practical reasons: it allowed the development of simple and efficient compilers. Lisp18, another programming language developed in the late 1950s for research in artificial intelligence, made the opposite choice: a program is just a particular interpretation of a data structure. Lisp programmers routinely assemble data structures and then execute them as programs. However, early Lisp implementations were slow compared to Fortran, and thus never became popular in computational science, with the notable exception of computer algebra systems.\n\nIn the early days of computational science, a theoretician would define a model with pencil and paper, and then write a program to do a specific computation based on that model, such as computing an integral or solving a differential equation numerically. The computation on the computer simply replaced the earlier practice of manual computation. Models were published in journal articles, just like in the pre-computing era. A computer program was considered an implementation of the model and testing the program involved comparing its output with results from analytical manipulation of the model for suitable input values.\n\nWith the rapid increase of computational power, scientists could handle ever more complex models, and in particular models far too complex to be managed with pencil and paper. But scientific publication remained in the pencil-and-paper world for a few more decades, because electronic communication became feasible only with the rise of the Internet in the 1990s. Scientists could thus work with computational models that were too complex for publication, and as a result they stopped publishing their models. With the separation of models and programs being discouraged by the computational tools, and in the absence of any motivation to formulate computational models independently from programs for communication, the fusion of models and programs became almost inevitable.\n\n\nSoftware as a notation for scientific knowledge\n\nIn the previous section, I have explained the undesirable consequences of the fact that computational models are often inseparably intertwined with the software tools that work on them. There is another important problem resulting from the fusion of tools and models, which is related to the different time scales on which science and computing technology evolve at the moment. This problem could disappear in the unlikely case that progress in computing technology slows down in the future, but it currently requires immediate attention if we want to preserve the scientific heritage of the last decades.\n\nKnowledge has a finite lifetime. Even if information storage media could be preserved forever, the meaning of the information they contain is ultimately lost because the semantic context in which it was encoded cannot be recorded. The best examples are historical written documents that nobody can read today, because the languages and writing systems used at the time have disappeared19.\n\nWritten human languages are the most stable semantic contexts we have: they change on a time scale of centuries to millennia. Scientific jargon and scientific notations are even more short-lived. Journal articles written 100 years ago are difficult to understand for today’s scientists. The original writings of Galileo or Newton can be understood only by scholars specialized in the history of science. The time scale on which original publications remain understandable is a few decades. This doesn’t mean of course that knowledge is lost rapidly. As the original writings become less and less clear, the aspects that are recognized as particularly important are constantly reformulated in review articles, monographs, and textbooks. This is why the insights of Galileo and Newton are still accessible to today’s physicists.\n\nSoftware as a notation for knowledge representation has a much shorter lifetime than scientific writing and mathematical notation. There are two approaches to understanding software: (1) studying it theoretically, by reading the source code and the documentation, and (2) observing its behavior, by running the program. Practice has shown that both approaches must be combined for a successful understanding of non-trivial software. Reading the source code permits making hypotheses about what the program does, which are then checked by running it on suitable input data. Source code remains intelligible as long as the language it is written in remains in active use. Depending on the language, this implies a time scale of a few years to at best one or two decades. Running a piece of non-trivial software without modifications is rarely possible after more than a few years. Software requires regular “maintenance” to remain usable. This maintenance consists in updating the source code and the installation procedures to adapt them to changes in the computing environment (compilers, operating systems, etc.) and in the dependencies (libraries, etc.). Maintenance is expensive and economically feasible only for widely used programs. Moreover, it generally proceeds in parallel with improvements in the models and methods implemented by the software. Today’s working version of a piece of scientific software does not necessarily reflect the models and methods that were implemented in its predecessor used a few years ago for an important computational study. Technical solutions such as the use of version control systems and archiving the exact code used for a specific scientific study can help to alleviate this problem, but they are not a panacea: they do not provide a code that works 30 years from now and implements today’s models and methods.\n\nThe consequence of the different time scales on which scientific knowledge and computing technology evolve is that we are losing scientific knowledge encoded in the form of software faster than it can be integrated into the reformulation process of science. For many computational studies performed during the last decades, it is already impossible to find the exact models and methods that were used. By applying the recommendations of the Reproducible Research movement, i.e. by publishing and archiving software and data, we can preserve the original expressions of this scientific knowledge, but not the semantic context.\n\nThis knowledge rot problem concerns not only models and methods that are embedded in scientific software, but also data stored in formats that are proprietary and thus defined by the software that reads and writes them. When the software becomes unusable, the data becomes unreadable. This aspect is much more widely recognized and there is a general consensus among experts in scientific data management that proprietary data formats are unsuitable for publishing and archiving purposes. The same attitude should be adopted with respect to models.\n\n\nShifting the focus from tools to models\n\nSolving the problems that I have discussed above would require most of all a shift of focus in computational science. Instead of concentrating on tools, which then subsume models and imprison data, we should focus on models and data as the primary items of interest for science. Before thinking about the question “How can I best do this computation?”, we must first consider the questions “What data and models does this computation depend on?” and “What will be the result of this computation?”\n\nSuch a shift of focus does not happen overnight. On the contrary, I would expect it to take many years, or maybe decades. In the following, I will outline some concrete steps to make it happen. First, I will discuss short-term actions that can be taken immediately and do not require profound changes to the scientific software and workflows that we use today. These actions will improve the understanding of the models implemented in scientific software, and will make it possible to discuss models in the scientific literature. I will then describe a second set of actions which require a serious research and development effort, but also offer significant benefits in return: the possibility to turn scientific models into first-class digital objects that can be published and archived, and the possibility to verify scientific software by formal proofs.\n\nThe main short-term action that must be taken is a thorough documentation of the scientific models that are implemented in a piece of software. Such documentation should explain the models in plain words and in mathematical notation, and point the reader to the relevant parts of the source code. Moreover, it should discuss how compilation and installation options and data in input files affect the models. As a guideline for deciding if a given feature belongs to the model or the tool, consider the interpretation of the results of the computation. Anything that changes these results in a way that must be understood for their interpretation is part of the model. Computational studies should cite the model documentation of the software that was used, and provide the values of all relevant compiler and installation options and input parameters.\n\nA related short-term action is writing reference implementations of scientific models in the form of programs optimized for clarity rather than performance or flexibility. Such reference implementations are at the same time a precise documentation of the model and executable programs whose results can be used to validate the results of more complex software written to be used as a tool. Writing reference implementations, like writing better documentation, takes time, and therefore one condition for making it happen is the creation of suitable incentives.\n\nFurther useful actions can be taken to improve scientific software without introducing any profound changes. User interfaces, a category which includes command-line options and the syntax of input files, can be redesigned to clearly separate model-related information (typically model parameters) from tool-related information. A clear distinction helps users to better understand the techniques they apply. Software developers can also aim for better modularity with respect to models: the source code of the program can be restructured to concentrate model-related aspects in as few source code files as possible.\n\nReaping the full benefits of a separation between models and tools requires more profound changes to the structure of scientific software. Models and tools must become distinct entities, which are developed, tested and published independently. Tools read in model specifications as input data. Such an approach is technically feasible today, due to the enormous progress that computer science has made since the 1950s. Domain-specific languages can be designed for the definition of scientific models, and translated by tools based on compiler technology into efficient code for today’s and tomorrow’s computers. A significant amount of research and development remains to be done, but it is justified by the improvements in the quality of computational science that it will make possible.\n\nThe most immediate benefit is that models will become well-defined citable entities. A model that has been specified in a formal machine-readable notation can be published and cited via a Digital Object Identifier (DOI). Tools can be written to define, explore, modify and evaluate models. In particular, tools that very similar in spirit to today’s computer algebra systems can be used to create approximations and combinations of scientific models. Theoreticians will be able to work with computable models in electronic form just like they used to work with mathematical models on paper in the past. Formal model specifications are not subject to the rapid evolution in computing technology, and can therefore be expected to be much more stable over time than today’s models embedded in software tools.\n\nFormalized models can also play an important role in future human interfaces to science, as used for communicating results and teaching students. It is foreseeable that static publications such as today’s articles will be replaced by dynamic and interactive presentation and visualization techniques (see 20 and 21 for examples). Creating such presentations on top of executable models ensures consistency between explanations and applications. With models being digital objects with clear semantics and a stable reference through a DOI, they become accessible to content mining and bibliometric analysis. It will be possible to compile databases of models used in published studies, which can then be annotated with validation information. The output of science shown in Figure 1 will become more formalized than it is now, which is likely to improve the quality of science overall.\n\nAnother important gain in reliability can be expected from software technology. The automatic program verification methods that are currently developed (see e.g. 22 for a non-trivial practical application) will become available for scientific software23. These approaches use automated proofs to verify that a program’s output conforms to its specification. These cannot be applied to today’s scientific software because it has no formal, and thus machine-readable, specification. The reason for this are the integrated models. Mathematical proof techniques cannot validate a model, because its validity is determined by comparison to observational data. However, given a formalized model, mathematical proof techniques can verify that a software tool correctly implements this model. This is probably the single most important element for improving trust in scientific software and thus computational science. However, to make this happen, a much closer collaboration of computational scientists and computer scientists would be required in the future.\n\nThe specificity of floating-point arithmetic deserve a special discussion, both because of its central role in much of scientific software and because of its reputation of being the source of intractable problems.\n\nFirst of all, it is worth pointing out that floating-point arithmetic can be defined as rigorously as integer arithmetic. The IEEE 754 standard24 provides a well-defined data representation at the bit level and a set of well-defined deterministic operations. Much of the mysterious behavior attributed to floating-point arithmetic is due to the fact that programmers and programming language designers reason about floating-point numbers as if they were real numbers, in particular assuming associativity for addition and multiplication. This happens partly by mistake (a lack of understanding of floating-point arithmetic), and partly out of the desire to create more opportunities for code optimization by compilers (see the discussion in 22).\n\nUnfortunately, none of the programming languages currently popular for scientific computing define the semantics of floating-point operations precisely enough to give the programmer a full control over the result of a calculation. As a consequence, the output of any program using floating-point arithmetic depends on choices made by compiler writers. Thus a scientific model specified with precise floating-point semantics cannot be implemented correctly using today’s scientific programming languages. This situation is in fact a consequence of the attitude that I have described in the introduction: computational science is so much focused on the performance of the computations and so little on the correctness of the results that there is no incentive for language designers and implementors to improve the situation.\n\nHowever, this does not mean that the actions I have described above are doomed to fail. The goal is to change the currently dominant attitudes. This should also lead to the development of programming tools that provide full control over floating-point operations. Moreover, it is not at all evident that floating-point numbers will continue to occupy a dominant role in scientific computing in the long run. Their popularity is mainly due to the at least apparent ease they offer for constructing computable approximations to the scientific models of the pre-computing era, which use real numbers to describe continuous physical quantities. It is well possible that other number representations will be used in the future. The recently proposed DEC64 format25, which aims to replace both integers and floating-point numbers, shows that there is still interest in improving number handling in computer software.\n\n\nRelated ideas and approaches\n\nThe problem that technical details tend to swamp the result-relevant aspects in program source code is not specific to scientific computing. Among the many software engineering approaches that aim to improve the situation, Model-Driven Engineering26 is the one most similar to the approach that I have outlined. It introduces the notion of a model as the specification of what a program is supposed to do. Program generators then produce an efficient implementation. However, like all of software engineering, Model-Driven Engineering has the goal of producing better tools. The models are little more than tool specifications, and are normally not accessible to the users of the finished software.\n\nSeveral scientific software packages are based on domain-specific languages (DSLs) that allow users to write down certain aspects of their problem in a notation that is more compact and familiar than a programming language. An example is the FEniCS package for solving differential equations27, whose DSL provides a means to write a differential equation in a notation that is close to traditional mathematics. However, the focus is on the mathematical equations rather than on the computable model, which consists of more than just the equations (boundary conditions, meshes, etc.). The distinction between the DSL and the implementation language is made for convenience of notation, not for a separation of concerns. This characteristic is shared by the other scientific DSLs that I am aware of. These DSLs have the goal of facilitating the technically most challenging part of encoding science in a computer program, but they do so clearly in the context of tool development.\n\nOrchard and Rice28 propose an “agenda for programming language research” in computational science which addresses many of the topics discussed here from the point of view of programming language research. Their agenda represents a continuation of the DSL approach described in the last paragraph. The authors insist particularly on the separation of concerns between scientific models and software implementation details, and propose a path of evolution for existing scientific software. Their article contains many references to prior work of interest.\n\nMurray-Rust and Murray-Rust’s “Reproducible Declaratron”29 proposes and implements ideas which are similar in many respects to what I have outlined in this article. Their approach is based on a long-term effort towards making scientific documents more precise and at the same time machine-readable by adding semantic markup. For example, in the plain-text sentence “the experiment was run at a temperature of 21 degrees”, the temperature specification would be replaced by XML elements indicating the type of quantity (temperature), the value (21), and the unit (degrees Celsius), with each part having a clearly defined meaning written down in a dictionary. The “Reproducible Declaratron” adds computation to this framework, applying the principle that formulae and algorithms are data. The authors do not make an explicit distinction between computable models and computational tools. They do make the distinction between a “formula” and a “computation”, which for the examples they discuss is very similar to the model-tool distinction, but is limited to models derived from mathematical equations. An outstanding feature of their approach is that it moves formulae and computations from computational tools into scientific publications.\n\nFinally, my own ActivePapers project30 provides a framework for computational science that does a first step in the direction I advocate in this paper: it shifts the focus from doing computations to publishing computational methods and results. An ActivePaper is a publishable and citable package of data sets, including executable codes as another kind of data. Every data set in a published ActivePaper has an automatically resolvable reference. Moreover, the framework was explicitly designed to include code transformation and code generation. However, suitable domain-specific model languages and tools that work on them remain to be developed.\n\n\nConclusion\n\nIn the preceding sections, I have explained that (1) the way we currently perform and publish computational science is unsatisfactory and that (2) we can and should improve our attitudes and technology. The situation I have described is a symptom of a lack of exchange between the natural sciences and research in computer science. Today’s computational scientists see computer science as an engineering discipline that provides them with ever increasing number crunching power. Their own training in computational techniques is usually limited to managing the practicalities of working with software tools. From the other side of the fence, computer scientists see scientific computing as almost synonymous with high-performance computing.\n\nIn the past centuries, much of the progress in science was due to an interplay between mathematics and physics in a domain of research now called “mathematical physics”. It was conducted by scientists who were at the same time application-oriented mathematicians and mathematically minded physicists. Science in the 21st century would benefit from a similar approach at the interface between computation and theoretical science. Computational scientists would discover that computers are not only convenient slaves to which they can offload laborious computations, but also tools that can improve our understanding of scientific models. Computer scientists would discover how computation plays a fundamental role in our efforts in understanding natural phenomena.", "appendix": "Competing interests\n\n\n\nThe author declares to 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\nThe author would like to thank Dominic Orchard for comments on a draft of this article.\n\n\nReferences\n\nMerali Z: Computational science: ...Error. Nature. 2010; 467(7317): 775–777. PubMed Abstract \| Publisher Full Text\n\nPeng RD: Reproducible research in computational science. Science. 2011; 334(6060): 1226–1227. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoppa LN, McInerny G, Harper R, et al.: Computational science. Troubling trends in scientific software use. Science. 2013; 340(6134): 814–815. PubMed Abstract \| Publisher Full Text\n\nStodden V: Reproducible research: Tools and strategies for scientific computing. Comput Sci Eng. 2012; 14(4): 11–12. Publisher Full Text\n\nWikipedia, Open science. — Wikipedia, the free encyclopedia. 2014. Reference Source\n\nGent I, Kotthoff L: Recomputation.org home page. Reference Source\n\nDrummond C: Replicability is not reproducibility: nor is it good science. 2009. Reference Source\n\nFrigg R, Hartmann S: Models in science. In The Stanford Encyclopedia of Philosophy (Fall 2012 Edition). E. N. Zalta, Ed. 2012. Reference Source\n\nHestenes D: Notes for a modeling theory. In Proceedings of the 2006 GIREP conference: Modelling in Physics and Physics Education, E. van den Berg, A. Ellermeijer, and O. Slooten, Eds. 2006. Reference Source\n\nHestenes D: Modeling games in the Newtonian world. Am J Phys. 1992; 60(8): 732–748. Publisher Full Text\n\nWikipedia, Computable function — Wikipedia, the free encyclopedia. 2014. Reference Source\n\nWikipedia, Three-body problem — Wikipedia, the free encyclopedia. 2014. Reference Source\n\nMarmaras B, Wang JJ: Simulation and visualization of few-body systems and the differential precession of Mercury. Comput Sci Eng. 2014; 16(1): 42–50. Publisher Full Text\n\nDowek G: Les métamorphoses du calcul: une étonnante histoire de mathématiques. Paris: Édition Le Pommier. 2007. Reference Source\n\nSussman GJ, Wisdom J: The role of programming in the formulation of ideas. MIT Artificial Intelligence Laboratory, Tech. Rep. AIM-2002–018. 2002. Reference Source\n\nMairson HG: Functional geometry and the traité de lutherie: Functional pearl. In Proceedings of the 18th ACM SIGPLAN International Conference on Functional Programming ser. ICFP ’13. New York, NY USA: ACM. 2013; 123–132. Reference Source\n\nBackus JW, Beeber RJ, Best S, et al.: The Fortran automatic coding system. In Papers Presented at the February 26–28, 1957, Western Joint Computer Conference: Techniques for Reliability, ser. IRE-AIEE-ACM ’57 (Western). New York, NY USA: ACM. 1957; 188–198. Publisher Full Text\n\nMcCarthy J: Recursive functions of symbolic expressions and their computation by machine, part I. Commun ACM. 1960; 3(4): 184–195. Reference Source\n\nWikipedia, Undeciphered writing systems. — Wikipedia, the free encyclopedia. 2014. Reference Source\n\nVictor B: Scientific communication as sequential art. 2011. Reference Source\n\nVictor B: Explorable explanations. 2011. Reference Source\n\nBoldo S, Jourdan JH, Leroy X, et al.: A formally-verified C compiler supporting floating-point arithmetic. In ARITH, 21st IEEE International Symposium on Computer Arithmetic. IEEE Computer Society Press. 2013; 107–115. Publisher Full Text\n\nIonescu C, Jansson P: Testing versus proving in climate impact research. In 18th International Workshop on Types for Proofs and Programs (TYPES 2011), ser. Leibniz International Proceedings in Informatics (LIPIcs), N. A. Danielsson and B. Nordström, Eds., vol. 19. Dagstuhl, Germany: Schloss Dagstuhl–Leibniz-Zentrum fuer Informatik. 2013; 41–54. Publisher Full Text\n\n754-2008 - IEEE standard for floating-point arithmetic. Publisher Full Text\n\nCrockford D: DEC64 number format. 2014. Reference Source\n\nSchmidt DC: Guest editor’s introduction: Model-driven engineering. Computer. 2006; 39(2): 25–31. Reference Source\n\nLogg A, Mardal KA, Wells G: Automated Solution of Differential Equations by the Finite Element Method. ser. Lecture Notes in Computational Science and Engineering. 2012; 84. Reference Source\n\nOrchard D, Rice A: A computational science agenda for programming language research. In International Conference on Computational Science, 2014. Reference Source\n\nMurray Rust P, Murray Rust D: Reproducible physical science and the declaratron. In Implementing Reproducible Research, V. Stodden, F. Leisch, and R. D. Peng, Eds. Chapman and Hall/CRC. 2014. Reference Source\n\nHinsen K: A data and code model for reproducible research and executable papers. Pro Comput Sci. 2011; 4: 579–588. Publisher Full Text" }	[ { "id": "4793", "date": "16 May 2014", "name": "David Hestenes", "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 exceptional paper in many ways. First, it is extremely well written and well organized. Second, it deals with an important, complex subject, providing a careful, novel analysis and parsing of terms. Though the subject cuts across many scientific domains, the author demonstrates clear competence in each. Finally, I think Hinsen is right in advocating primacy of models over tools (as he has so carefully defined them).", "responses": [] }, { "id": "4684", "date": "27 May 2014", "name": "Patrik Jansson", "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 position paper addresses the current state of computational science (the term is not defined, but appears to mean the use of computational models as an essential means of scientific investigation). As the author says, \"the central question in computational science is: why should we trust the results of a non-trivial computation?\". At the moment, the answer appears to be: \"we probably shouldn't\", clearly an unsatisfactory state of affairs.This is a point which has been made before, and the paper references a number of previous attempts to come to grips with it. The author's own approach is based on the introduction of an interesting conceptual distinction: between models and tools. It is then argued that many of the problems in today's computational science arise because of \"blurring the distinction\".That the distinction is not unproblematic is illustrated by the following difficulty. On page 4, tools are defined to be \"both physical objects and mathematical theorems or procedures (e.g. calculus or algebra)\". Later, we read: \"the conclusion of a scientific study needs to be independent of a specific tool to deserve the name 'scientific'\". On the previous page, however, we have the example of a description of the solar system as a set of differential equations. This seems to be crucially dependent on calculus. Is it therefore not scientific?As this example shows, the paper is sure to provide food for thought and for debate. We find that it does contribute to highlighting and clarifying the problems we encounter in computational science, and as such it deserves to be published and to reach a wider audience.There is one error which must be corrected before that, however: namely, on page 3, where the functions \"sin\" and \"cos\" are said to be uncomputable, by virtue of being transcendent. In fact, it is well known that they are computable (as are all functions that can be expressed by power series, which is one of the examples in Turing's 1936 paper). If the author has a different notion of computability in mind, then it should be stated explicitly. (See the labels [Approx0], [Approx1] and [Approx2] below for some more details.)Another somewhat strange statement (but not an error) is that \"heuristic methods that involve random choices [...] do not produce a unique result for a given input and therefore do not qualify as scientific models\". One would think that they do, in fact, produce a unique result: namely a probability distribution, and that computer simulations can be used to sample this distribution to whatever precision necessary.There are, additionally, a number of infelicities which might be addressed before publication, such anthropomorphic expressions (\"computers help with collecting..., evaluating, ..., and communicating\" instead of \"computers are used for collecting ..., evaluating, ..., and communicating\"), or the less-than-helpful figure 1, which does not seem to bring anything over its one line description in the caption.Some detailed comments:Abstract: \"As a consequence, these crucial pieces of information have disappeared from the scientific record.\" We suggest some reformulation. It is not that they have first been there and later \"disappeared\" but rather they were absent all along and now (with more complexity developing) they are increasingly difficult (or impossible) to generate from what is in \"the scientific record\". [Approx0] \"But this solution contains transcendental functions (sines and cosines), which are not computable and therefore must be replaced by computable approximations, e.g. power-series expansions.\" Reformulate: perhaps move the text from [Approx2] up and reformulate. \"... there are known algorithms that can find ... in finite time to any specified precision.\" [Approx1] \"they cannot make predictions that are immediately comparable to observed data\": All observed data is of finite precision so it can be claimed that comparing model results to\n\nobserved data is often (but not always) feasible. \"The very fact that a program runs and produces results proves that\n\nthe model specification is complete and unambiguous, ...\" The semantics of some languages (like C) is ambiguous (non-deterministic). \"... the importance of correctly identifying, understanding and documenting approximations.\" We strongly agree with this part and think it would be worth pointing to interval arithmetic as one way to handle this problem. [Approx2] \"It is even a computable model in the sense of computability theory, in that there are known algorithms that can find the global minimum in finite time to any specified precision.\"\n\nNote that this looks inconsistent within the paper: earlier (at [Approx0]) you claim that sin and cos are not computable, but with this definition they are. It would be preferable to skip \"of course\" here: \"This doesn’t mean of course that knowledge is lost rapidly.\" := \"This doesn’t mean that knowledge is lost rapidly.\" \"tools that very similar in spirit\" := \"tools that are very similar in spirit\".", "responses": [ { "c_id": "851", "date": "05 Jun 2014", "name": "Konrad Hinsen", "role": "Author Response", "response": "The referees raise some interesting points, which I hope to have addressed in the revised version of my article.Concerning the use of calculus in defining mathematical models through differential equations, it is important to distinguish between axioms and definitions on one hand, and theorems and their applications on the other hand. I have added a short sentence to make this distinction explicit. Theorems and procedures are tools, and scientific models should indeed not depend on them. Axioms and definitions are simply the semantic context for mathematical equations, i.e. the language in which mathematical models are written. Newton's equations are based on the concept of a derivative, but do not rely on any particular theorem or technique from calculus.The referees are right about the computability of sines and cosines. I have corrected this statement in the revision.I have also added a paragraph at the end of the \"Tools and methods\" section to clarify the distinction between probabilistic models and heuristics involving arbitrary or random choices, which wasn't as carefully explained as it should have been.Minor points that I have addressed in the revision are:the ambiguity of languages like Cinterval arithmetic as an alternative to floating-point arithmetictypos and stylistic issues" } ] } ]	1	https://f1000research.com/articles/3-101
https://f1000research.com/articles/3-23/v1	23 Jan 14	{ "type": "Short Research Article", "title": "Fiberoptic monitoring of central venous oxygen saturation (PediaSat) in small children undergoing cardiac surgery: continuous is not continuous", "authors": [ "Francesca G. Iodice", "Zaccaria Ricci", "Roberta Haiberger", "Isabella Favia", "Paola Cogo", "Francesca G. Iodice", "Roberta Haiberger", "Isabella Favia", "Paola Cogo" ], "abstract": "Background: Monitoring of superior vena cava saturation (ScvO2) has become routine in the management of pediatric patients undergoing cardiac surgery. The objective of our study was to evaluate the correlation between continuous ScvO2 by the application of a fiber-optic oximetry catheter (PediaSat) and intermittent ScvO2 by using standard blood gas measurements. These results were compared to those obtained by cerebral near infrared spectroscopy (cNIRS).Setting: Tertiary pediatric cardiac intensive care unit (PCICU).Methods and main results: A retrospective study was conducted in consecutive patients who were monitored with a 4.5 or 5.5 F PediaSat catheter into the right internal jugular vein. An in vivo calibration was performed once the patient was transferred to the PCICU and re-calibration took place every 24 hours thereafter. Each patient had a NIRS placed on the forehead. Saturations were collected every 4 hours until extubation. Ten patients with a median age of 2.2 (0.13-8.5) years and a weight of 12.4 (3.9-24) kg were enrolled. Median sampling time was 32 (19-44) hours: 64 pairs of PediaSat and ScVO2 saturations showed a poor correlation (r=0.62, 95% CI 44-75; p<0.0001) and an average difference of -0.38 with a standard deviation of 13 and 95% limits of agreement from -26 to 25. Thirty-six pairs of cNIRS and ScVO2 saturations showed a fair correlation (r=0.79, 95% CI 0.60-0.89; p<0.0001) an average difference of -1.3 with a standard deviation of 7 and 95% limits of agreement from -15 to 12. Analysis of median percentage differences between PediaSat and ScvO2 saturation over time revealed that, although not statistically significant, the change in percentage saturation differences was clinically relevant after the 8th hour from calibration (from -100 to +100%).Conclusion: PediaSat catheters showed unreliable performance in our cohort. It should be further investigated whether repeating calibrations every 8 hours may improve the accuracy of this system. CNIRS may provide similar results with a lower invasiveness.", "keywords": [ "Postoperative pediatric patients who have undergone cardiac surgery may benefit from venous saturation monitoring to assess oxygen delivery and as an indirect method of systemic perfusion1", "2. A true mixed venous sample (SvO2) is drawn from the tip of the pulmonary artery catheter and includes all of the venous blood returning from the head and arms (via superior vena cava)", "the gut", "the kidneys and lower extremities (via the inferior vena cava) and the heart (via the coronary sinus). In recent years", "however", "superior vena cava oxygen venous saturation (ScvO2) has replaced mixed venous saturation in many clinical settings and it is considered a reliable surrogate of SvO23. In neonates and pediatric patients", "the placement of a pulmonary artery catheter is not routine and may be problematic. In these patients ScvO2 monitoring plays an important therapeutic role4. ScvO2 monitoring can be done intermittently (by the “traditional” co-oximetry method with serial blood withdrawals) or by reflectance oximetry through a fiber-optic catheter4", "5. A new dedicated multilumen (PediaSat", "Edwards Lifesciences", "Irvine", "CA", "USA) central venous catheter (CVC) with incorporated fiber-optic technology for continuous oxygen saturation monitoring has been designed for use in neonates and pediatric patients. If correctly placed with the distal tip in the superior vena cava", "this catheter is able to provide a reliable on-line measurement of the ScvO24", "6. There is evidence that continuous monitoring of ScvO2 may have beneficial effects in the resuscitation of septic patients and after complex congenital heart disease7", "8. According to manufacturer’s recommendations", "after the first calibration the PediaSat catheter should provide consistent information for the following 24 hours. After this a new calibration is recommended in order to correct for potential drift from the true value9." ], "content": "Introduction\n\nPostoperative pediatric patients who have undergone cardiac surgery may benefit from venous saturation monitoring to assess oxygen delivery and as an indirect method of systemic perfusion1,2. A true mixed venous sample (SvO2) is drawn from the tip of the pulmonary artery catheter and includes all of the venous blood returning from the head and arms (via superior vena cava), the gut, the kidneys and lower extremities (via the inferior vena cava) and the heart (via the coronary sinus). In recent years, however, superior vena cava oxygen venous saturation (ScvO2) has replaced mixed venous saturation in many clinical settings and it is considered a reliable surrogate of SvO23. In neonates and pediatric patients, the placement of a pulmonary artery catheter is not routine and may be problematic. In these patients ScvO2 monitoring plays an important therapeutic role4. ScvO2 monitoring can be done intermittently (by the “traditional” co-oximetry method with serial blood withdrawals) or by reflectance oximetry through a fiber-optic catheter4,5. A new dedicated multilumen (PediaSat; Edwards Lifesciences, Irvine, CA, USA) central venous catheter (CVC) with incorporated fiber-optic technology for continuous oxygen saturation monitoring has been designed for use in neonates and pediatric patients. If correctly placed with the distal tip in the superior vena cava, this catheter is able to provide a reliable on-line measurement of the ScvO24,6. There is evidence that continuous monitoring of ScvO2 may have beneficial effects in the resuscitation of septic patients and after complex congenital heart disease7,8. According to manufacturer’s recommendations, after the first calibration the PediaSat catheter should provide consistent information for the following 24 hours. After this a new calibration is recommended in order to correct for potential drift from the true value9.\n\nNear-infrared spectroscopy (NIRS) is a noninvasive technique that measures continuous regional tissue oxygenation and is routinely used during pediatric cardiac surgery. NIRS measures the percentage oxygenated hemoglobin level in tissue beds. It is commonly used to determine cerebral tissue oxygen saturation (cerebral NIRS or cNIRS) and renal somatic tissue oxygen saturation (renal NIRS). cNIRS has been proposed to estimate adequacy of oxygen delivery due to a significant correlation existing between ScvO2 and cerebral NIRS10–13.\n\nThe objective of our study was to evaluate the correlation between continuous ScvO2 and intermittent SvO2 by using standard blood gas measurements during specific time points in postoperative pediatric cardiac patients. In particular we aimed to verify when calibration of a continuous ScvO2 catheter should be repeated in order to optimize the detection of potentially significant differences between continuous ScvO2 and intermittent ScvO2. Furthermore we also evaluated the correlation between cNIRS and intermittent SvO2.\n\n\nMaterials and methods\n\nA retrospective observational study was conducted in a tertiary pediatric cardiac intensive care unit from January 2013 until May 2013. Children undergoing elective cardiac surgery for congenital heart disease, and who had a PediaSat catheter placed (the indication was given by the attending anesthesiologist) at the Bambino Gesù Children’s Hospital, Rome, Italy, were enrolled in the study. Inclusion criteria were: 1) patient had been scheduled for elective cardiac surgery with cardiopulmonary bypass; 2) the patient was clinically indicated for central venous catheter placement with a size of 4.5 to 5.5 F; 3) the patient’s age was within the selected range: newborn >38 weeks gestation to child <10 years old; 4) the patient’s weight was >3.0 kg. Exclusion criteria were: 1) emergency operation; 2) need for or decision to position a femoral central venous catheter.\n\nThe INVOS 5100C Cerebral Oximeter (Somanetics, Troy, MI, USA) was used in all patients undergoing pediatric cardiac procedures with cardiopulmonary bypass according to the institutional protocol.\n\nAfter induction of anesthesia, tracheal intubation, and radial or femoral artery cannulation, a central venous PediaSat catheter was inserted into the superior vena cava through the right internal jugular vein. The catheter was inserted using ultrasound guidance. Correct positioning was confirmed via transesophageal echocardiography and by a post-operative chest x-ray. The size and length of the catheters were decided on the basis of the patients’ weight. After catheter insertion an in vivo calibration was performed and repeated once the patient was transferred to the cardiac intensive care unit (CICU). The catheter was re-calibrated every 24 hours thereafter. Following induction of anesthesia, a NIRS sensor was placed on the patient’s forehead after adequate scrubbing of the skin.\n\nData were retrieved from the institutional database and missing data were acquired from patients’ clinical charts. Demographics, surgical procedure, cardiac bypass time and cross clamp time were recorded. In order to obtain ScvO2, blood was withdrawn from the distal port of the PediaSat catheter. Blood gases were analyzed in heparinized 1 ml syringes, within 60 seconds from withdrawal, with the GEM4000 blood gas analyzer (Brennan & Company, Dublin, Ireland). For each ScvO2 value, a PediaSat and cNIRS saturation was collected. According to the institutional protocol, patients’ oxygen saturations are reported into clinical chart every 4 hours until extubation (the first value being reported 1 hour after calibration). After patient extubation, data collection was terminated.\n\nA Spearman test was chosen for correlation estimation, and a Bland Altman analysis was used to verify bias and agreement of correlated variables. Results are expressed as median (interquartile range). Wilcoxon signed rank test was used for paired group comparison. One-way analysis of variance (Kruskal-Wallis non parametric test) was applied in order to evaluate difference of saturations as measured by the PediaSat and ScvO2 over time. Agreement between the two methods for tracking changes in SvO2 was quantified using polar plots: acceptable calibration was defined as an angular mean bias of less than ±5° and the percentage of data points lying within radial limits of ±30° from the polar axis was assessed14. A P value <0.05 was considered significant. Statistical analysis was performed with the GRAPHPAD PRISM 5.0 software package (GraphPad Software, San Diego, CA, USA).\n\nInstitutional review board (“Comitato Etico per la Sperimentazione Clinica”) approved the study and waived the need for parental informed consent due to the retrospective nature of the study.\n\n\nResults\n\nTen patients with a median age of 2.2 (0.13–8.5) years and a weight of 12.4 (3.9–24) kg were enrolled. Median mechanical ventilation duration was 36 hours (12–48). Cardiologic diagnoses and surgical procedures with cardiopulmonary bypass (CPB) details are reported in Table 1. Median sampling time was 32 (19–44) hours: at 48 hours all patients were extubated and the data collection never lasted longer than this. Sixty-four pairs of PediaSat and ScvO2 saturations were available (Figure 1A): 5 missing pairs are acknowledged (on the clinical chart, once was ScvO2 value alone reported and 4 times no value was reported). Median PediaSat venous saturation was 71 (64–81)% whereas the median ScvO2 value was 74 (64–82) (p=0.347). Correlation between these two methods was poor (r=0.62, 95% CI 44–75; p<0.0001) (Figure 2A). Bland Altman analysis revealed a bias of -0.38 with a standard deviation of 13 and 95% limits of agreement from -26 to 25 (Figure 2B). Thirty-six pairs of cNIRS and ScVO2 saturations were also available for analysis (Figure 1B): median cNIRS venous saturation was 74 (62–78)% whereas median ScvO2 values were 71 (62–78) (p=0.150). Correlation between these two methods was fair (r=0.79, 95% CI 0.60–0.89; p<0.0001) (Figure 2C). Bland Altman analysis revealed a bias of -1.3 with a standard deviation of 7 and 95% limits of agreement from -15 to 12 (Figure 2D). Analysis of median percentage differences between PediaSat and ScvO2 saturation over time revealed that there was not a significant modification over the 6 time points in the 24 hour timeframes (p=0.28) (Figure 3A). However, although not statistically significant, the change of percentage saturation differences was clinically relevant especially after the 8th hour after calibration when errors from -100% to +100% were noted both as under and overestimation by the PediaSat method (Table 2). Similar results were found when percentage cNIRS-ScvO2 differences were evaluated (p=0.86) (Figure 3B), although the error never exceeded 20% at any time point and cNIRS showed a slight tendency to systematic underestimation of true values (Table 2). Trending ability of PediaSat as assessed by a Polar plot showed a mean angular deviation from the polar axis of 90°, with 50% of the data points lying outside the radial limits of ±30° from the polar axis. Trending ability of cNIRS assessed by Polar plot showed a mean angular deviation from the polar axis of 42° with 30% of the data points lying outside the radial limits of ±30° from the polar axis.\n\nBSA: body surface area. LOMV: length of mechanical ventilation. CPB: cardiopulmonary bypass. Xclamp: cross clamp.\n\nSimultaneous ScvO2 values provided by intermittent blood gas analysis (BGA), the PediaSat catheter (A) and cerebral Near Infrared Spectroscopy (cNIRS) (B).\n\nA: Bland Altman plot of ScvO2 values obtained by blood gas analysis (BGA) and PediaSat. B: correlation of ScvO2 values obtained by BGA and PediaSat. C: Bland Altman plot of ScvO2 values obtained by BGA and cerebral near infrared spectroscopy (cNIRS). D: correlation of ScvO2 values obtained by BGA and cNIRS.\n\nPercentage difference between ScvO2 values obtained by blood gas analysis (BGA) and PediaSat (l A), and BGA and cerebral near infrared spectroscopy (cNIRS) (B), expressed over time.\n\n\n\n\nDiscussion\n\nEvaluation of oxygen delivery is optimized by measurement of continuous superior vena cava oxygen saturation. This is due to the fact that unexpected/sudden low cardiac output events may occur in the timeframe between serial ScvO2 evaluations; of note, the trigger for occasional venous samplings (hypotension, need for increasing vasoactive drugs dose, etc.) may also be prompted untimely with respect to cardiac decompensation. The ideal device to perform this continuous monitoring should be accurate and minimally invasive, it should provide at least two saturation values per minute, it should be applicable to smaller patients and, finally, it should be cost-effective. To date, such a device is not yet available in routine clinical practices. The PediaSat catheter does certainly have some of these features and several initial reports seemed to provide encouraging results in clinical practice2,4,5. Unfortunately, the application of PediaSat to our small cohort of patients below 10 kg of body weight provided unsatisfactory results: the PediaSat catheter provided unacceptable saturation differences, with respect to the reference ScvO2 values (-26 to 25%). Furthermore, such deviation from actual values did not show a systematic error on the device and the PediaSat did not display a consistent over- or underestimation of true ScvO2 values: hence adjustments did not seem possible in order to correct the measures. In terms of trend estimation, PediaSat provided fair results, although still far from being adequate in terms of reliable routine utilization. The reasons for its lack of accuracy are not completely clear, but they could be caused by imperfect functioning of the miniaturized technology. It is also possible that more frequent calibrations should be performed: in our cohort, precision of PediaSat started to decrease, in a clinically significant way, after the 8th hour after calibration and this trend was similar after each calibration (as seen in the subgroup of patients who passed the 24th monitoring hour and whose PediaSat catheter was therefore re-calibrated). On the other hand, cNIRS, whose application is favored due to its low invasiveness, showed similar results, if not slightly better, in terms of coupling with ScvO2 values, limits of agreement at Bland Altman analysis, drift from actual values over time and trending ability. It must be remarked, however, that cNIRS values may be significantly affected by ventilation, sedation, cardiac anatomy and temporal distance from the surgical procedure13.\n\nOur results conflict with some reports2,4,5 and agree with others15. However, this is one of the first studies evaluating PediaSat in routine practice, and not during a specifically designed study. Furthermore, we hypothesize that PediaSat catheter might improve its performance with the calibration repeated every 8 hours.\n\nOur study is certainly limited by the small subgroup of patients. However, more than 60 pairs of ScVO2 were available for comparison and it is likely that they were adequate for PediaSat catheter evaluation. The PediaSat catheter was not used for longer than 48 hours in this study and it would have been interesting to evaluate its performance over the third and subsequent calibrations. Finally, we acknowledge that the comparison between PediaSat and cNIRS monitor may be biased by the fact that cNIRS values were fewer in number and this may have randomly reduced the possibility for cNIRS errors. For these reasons we cannot definitely state which of the two monitoring techniques performs better.\n\nIn conclusion, 4.5 and 5.5 F PediaSat catheters showed unreliable performance during the early post operative course of children below 10 kg, who underwent cardiac surgery with CPB. It should be further investigated if repeating calibrations every 8 hours instead of every 24 hours may improve the accuracy of this system. At the moment cNIRS provides similar results with a lower invasiveness.\n\n\nData availability\n\nfigshare: Post-operative clinical data of children enrolled for comparison of ScvO2% with PediaSat and with cerebral NIRS values. doi: http://dx.doi.org/10.6084/m9.figshare.90091516", "appendix": "Author contributions\n\n\n\nFG Iodice and Z Ricci conceived and drafted the paper, R Haiberger and I Favia collected the data. P Cogo supervised the final version of the manuscript. All the authors agreed with 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 PediaSat catheters and the cNIRS sensors used in this data collection were provided by the Bambino Gesù Hospital.\n\n\nReferences\n\nReinhart K, Kuhn HJ, Hartog C, et al.: Continuous central venous and pulmonary artery oxygen saturation monitoring in the critically ill. Intensive Care Med. 2004; 30(8): 1572–1578. PubMed Abstract \| Publisher Full Text\n\nSpenceley N, Krahn G, Skippen P, et al.: Evaluation of a pediatric central venous oximetry catheter in critically ill children. Pediatric Crit Care Med. 2010; 11(1): 26–30. PubMed Abstract \| Publisher Full Text\n\nShepherd SJ, Pearse RM: Role of Central and Mixed Venous Oxygen Saturation Measurement in Perioperative Care. Anesthesiology. 2009; 111(3): 649–56. PubMed Abstract \| Publisher Full Text\n\nRanucci M, Isgrò G, De La Torre T, et al.: Continuous Monitoring of Central Venous Oxygen Saturation (PediaSat) in Pediatric Patients Undergoing Cardiac Surgery: A Validation Study of a New Technology. J Cardiothorac Vasc Anesth. 2008; 22(6): 847–852. PubMed Abstract \| Publisher Full Text\n\nMohnseni-Bod H, Frndova H, Gaitaro R, et al.: Evaluation of a new pediatric continuous oximetry catheter. Pediatr Crit Care Med. 2011; 12(4): 437–441. PubMed Abstract \| Publisher Full Text\n\nRanucci M, Isgro’ G, De La Torre T, et al.: Near-infrared spectroscopy correlates with continuous superior vena cava oxygen saturation in pediatric cardiac surgery patients. Paediatr Anaesth. 2008; 18(12): 1163–9. PubMed Abstract \| Publisher Full Text\n\nOtero RM, Nguyen HB, Huang DT: Early goal directed therapy in severe sepsis and septic shock revisited: Concepts, controversies and contemporary findings. Chest. 2006; 130(5): 1579–1595. PubMed Abstract \| Publisher Full Text\n\nBradley SM, Atz AM: Postoperative management: The role of mixed venous oxygen saturation monitoring. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu. 2005; 8(1): 22–27. PubMed Abstract \| Publisher Full Text\n\nEdwards PedaiSat catheter online brochure. Reference Source\n\nMarimón GA, Dockery WK, Sheridan MJ, et al.: Near-infrared spectroscopy cerebral and somatic (renal) oxygen saturation correlation to continuous venous oxygen saturation via intravenous oximetry catheter. J Crit Care. 2012; 27(3): 314.e13–314.e18. PubMed Abstract \| Publisher Full Text\n\nTortoriello TA, Stayer SA, Mott AR, et al.: A noninvasive estimation of mixed venous oxygen saturation using near-infrared spectroscopy by cerebral oximetry in pediatric cardiac surgery patients. Paediatr Anaesth. 2005; 15(6): 495–503. PubMed Abstract \| Publisher Full Text\n\nWeiss M, Dullenkopf A, Kolarova A, et al.: Near-infrared spectroscopic cerebral oxygenation reading in neonates and infants associated with central venous oxygen saturation. Paediatr Anaesth. 2005; 15(2): 102–109. PubMed Abstract \| Publisher Full Text\n\nRicci Z, Garisto C, Favia I, et al.: Cerebral NIRS as a marker of superior vena cava oxygen saturation in neonates with congenital heart disease. Paediatr Anaesth. 2010; 20(11): 1040–5. PubMed Abstract \| Publisher Full Text\n\nCritchley LA, Lee A, Ho AM: A critical review of the ability of continuous cardiac output monitors to measure trends in cardiac output. Anesth Analg. 2010; 111(5): 1180–92. PubMed Abstract \| Publisher Full Text\n\nBaulig W, Béttex D, Bürki C, et al.: The PediaSat continuous central SvO2 monitoring system does not reliably indicate state or course of central venous oxygenation. Eur J Anaesthesiol. 2010; 27(8): 720–5. PubMed Abstract \| Publisher Full Text\n\nIodice FG, Ricci Z, Haiberger R, et al.: Post-operative clinical data of children enrolled for comparison of ScvO2% with PediaSat and with cerebral NIRS values F1000Research. figshare. 2014. Data Source" }	[ { "id": "3896", "date": "27 Feb 2014", "name": "Sabino Scolletta", "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, Dr. Iodice and associates presented the results on the correlation between continuous and intermittent ScvO2 values, obtained with a fiber-optic oximetry catheter (PediaSat) and standard blood gas measurements, respectively. In addition, ScvO2 values were compared to those obtained by cerebral near infrared spectroscopy (cNIRS). The study was conducted on ten children undergoing cardiac surgery.Although this is a retrospective observational study, it has been well described. Also, the article is clear, and the results are well presented. The discussion is very concise but highlights sufficiently some positive and negative aspects of the devices under study. Furthermore, it makes comparisons with other studies showing similar or opposite results on this field. My major criticism is about statistical analysis. Indeed, as the measurements of oxygen saturation have been performed at different times, the adjustment for the effects of repeated measurements in the Bland- Altman analysis had to be used. Conversely, the authors have used standard Bland-Altman and correlation analyses.Despite the aforementioned criticism, I think the topic of this article is of interest. Indeed, monitoring tissue oxygenation in critically ill patients is crucial during cardiac surgery. In particular, children undergoing cardiac operations may benefit from assessing the adequacy of their oxygen delivery (DO2) with respect to their oxygen requirements. To this purpose, an indirect index of tissue perfusion is the mixed venous oxygen saturation (SvO2). However, to measure this parameter a pulmonary artery catheter (PAC) is required. Unfortunately, PAC is invasive and usually not practicable in children. Over the past years, the central venous oxygen saturation (ScvO2) has been used in many clinical settings as a surrogate of SvO2. Thus, ScvO2 can be measured continuously (fiber-optic catheter) or intermittently (blood withdrawals) and it is currently considered a possible alternative to SvO2. Besides, the measure of ScvO2 requires a less invasive technique, as for its measurement a central venous catheter is needed instead of a PAC.In this study, the main result was that the fiber-optic catheter PediaSat showed unreliable performance. However, the authors had repeated measures from a small cohort of children with <10 kg of body weight. This may have biased their findings. Further studies are warranted, as conflicting results on this issue still remain in the current literature.Of note, a key message from this investigation is that monitoring brain function with cNIRS is safe, continuous and easy to achieve in children undergoing cardiac surgery. In addition, this method has the main advantage of providing useful information on brain function, which can be considered a subtle marker of systemic hypoperfusion in case of low cerebral oxygen saturation. Monitoring cerebral function with cNIRS may avoid potential neurological complications, which are associated with increased morbidity and mortality, and poor quality of life.", "responses": [ { "c_id": "723", "date": "01 Mar 2014", "name": "Zaccaria Ricci", "role": "Author Response", "response": "We are thankful to the reviewer for the precious suggestions and kind comments. We agree that the topic of assessing the adequacy of children's oxygen delivery (DO2) with respect to their oxygen requirements is crucial in pediatric intensive care and we are convinced that this topic has not been addressed effectively, so far. Indeed he is right: Bland Altman analysis for repeated measures should have been performed for our data analysis. This will be certainly modified in the revised version of the manuscript. Of note we already performed the corrected analysis and did not found significantly different results. In light of this we think that the overall message of the study has not been affected by our methodological mistake." } ] }, { "id": "3867", "date": "04 Apr 2014", "name": "Mirela Bojan", "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 with pleasure the article by Dr Ricci and colleagues. The topic is of high interest, since SvO2 measurements represent standard monitoring in the postoperative management of several pediatric cardiac procedures. Also, ScvO2 is nowadays an accepted surrogate measurement of SvO2. Significant differences have been reported between continuous fiberoptic measurements and intermittent measurements in adults with cardiac surgery (Baulig et al., 2008), requiring repeated calibration of the continuous measurement instrument. The findings here suggest a requirement for calibration every 8hrs, albeit the population is too small to allow for conclusions.I believe there is a bias in the statistical analysis, since correlations have been estimated between repeated measurements – as a matter of fact, there is a temporal autocorrelation between the results of a same patient at different time points. However, this bias should result in overestimation of the relationship. The negative result concerning ScvO2-BGA correlation is somewhat “reinforced” by this bias. But one can question about the positive NIRS-BGA correlation. I understand that the population is too small to calculate correlations at each time point, or to analyse data using a mixed model for repeated measurement. Therefore, I suggest presenting the results in Figure 1 and Figure 3 separately for each patient, showing on the same graph the BGA, ScvO2 and NIRS measurements, and allow for visual assessment of the relationship in each patient at each time point. This should reassure the reader about the potential bias due to temporal autocorrelation. Since the correlation between ScvO2-BGA is negative, and since the authors do not demonstrate but only suggest a positive correlation between NIRS-BGA, this should be acceptable.", "responses": [ { "c_id": "755", "date": "05 Apr 2014", "name": "Zaccaria Ricci", "role": "Author Response", "response": "We wish to thank Dr Bojan for her valued suggestions; these are in line with the comments of Sabino Scolletta. As recommended, we will change the figures as per improved visualization of data and leave the remaining results unchanged. As remarked by the reviewer, the new figures will reassure the reader, in order to cope for repeated measure analysis, although confirming the results presented in the text." } ] } ]	1	https://f1000research.com/articles/3-23
https://f1000research.com/articles/3-125/v1	10 Jun 14	{ "type": "Research Article", "title": "Vitellogenin family gene expression does not increase Drosophila lifespan or fecundity", "authors": [ "Yingxue Ren", "Kimberly A. Hughes", "Kimberly A. Hughes" ], "abstract": "One of the most striking patterns in comparative biology is the negative correlation between lifespan and fecundity observed in comparisons among species. This pattern is consistent with the idea that organisms need to allocate a fixed energy budget among competing demands of growth, development, reproduction and somatic maintenance. However, exceptions to this pattern have been observed in many social insects, including ants, bees, and termites. In honey bees (Apis mellifera), Vitellogenin (Vg), a yolk protein precursor, has been implicated in mediating the long lifespan and high fecundity of queen bees. To determine if Vg-like proteins can regulate lifespan in insects generally, we examined the effects of expression of Apis Vg and Drosophila CG31150 (a Vg-like gene recently identified as cv-d) on Drosophila melanogaster lifespan and fecundity using the RU486-inducible GeneSwitch system. For all genotypes tested, overexpression of Vg and CG31150 decreased Drosophila lifespan and did not affect total or age-specific fecundity. We also detected an apparent effect of the GeneSwitch system itself, wherein RU486 exposure (or the GAL4 expression it induces) led to a significant increase in longevity and decrease in fecundity in our fly strains. This result is consistent with the pattern reported in a recent meta-analysis of Drosophila aging studies, where transgenic constructs of the UAS/GAL4 expression system that should have no effect (e.g. an uninduced GeneSwitch) significantly extended lifespan in some genetic backgrounds. Our results suggest that Vg-family genes are not major regulators of Drosophila life history traits, and highlight the importance of using appropriate controls in aging studies.", "keywords": [ "Aging (senescence) is an almost universal process in multicellular organisms", "in which organismal function and performance decline with age1", "2. Decreasing fertility and increasing mortality are general hallmarks of aging that are typically accompanied by a declining activity level", "altered metabolic rate", "and a higher susceptibility to predation", "parasites and disease3–5. Despite its ubiquity", "patterns and rates of aging vary enormously among", "and within species. A large body of theory and experiment explores the evolutionary causes of this diversity", "however", "the underlying molecular mechanisms are still poorly understood6", "7." ], "content": "Introduction\n\nAging (senescence) is an almost universal process in multicellular organisms, in which organismal function and performance decline with age1,2. Decreasing fertility and increasing mortality are general hallmarks of aging that are typically accompanied by a declining activity level, altered metabolic rate, and a higher susceptibility to predation, parasites and disease3–5. Despite its ubiquity, patterns and rates of aging vary enormously among, and within species. A large body of theory and experiment explores the evolutionary causes of this diversity; however, the underlying molecular mechanisms are still poorly understood6,7.\n\nEvolutionary life history (LH) theory generally assumes that allocation of energy among the competing demands of growth, development, reproduction and somatic maintenance lead to functional trade-offs among these processes8,9. Consistent with this idea of resource allocation trade-offs, lifespan and fecundity are generally negatively correlated in comparisons among species8,10, A prominent exception to this pattern occurs in many social insects. In many ants, termites, and bees11, reproductive females are both long-lived and highly fecund relative to other species. For example, queen black garden ants (Lasius niger) can live for at least 28 years, while laying hundreds of eggs per day12, and queen honey bees (Apis mellifera) have a maximum lifespan of 3–5 years while laying thousands of eggs per day13. In contrast, most non-social insects have adult longevity of less than one year, and have lower fecundity than social insect queens14.\n\nInvestigations in honey bees suggest that the Vitellogenin gene (Vg) produces a yolk protein precursor that is synthesized in the abdominal fat body and acts as an antioxidant and promotes longevity in queen bees15. Differential expression of Vg has also been associated with the differences in lifespan between different kinds of worker bees: higher expression is seen in \"winter bees\" which have a lifespan of 10 months to 1 year, and lower expression in \"summer bees\" with a lifespan of 30–50 days16–18. RNAi knockdown of Vg expression in workers resulted in lower oxidative stress resistance19 and shorter lifespan20. In contrast, RNAi knockdown of Vitellogenin-encoding genes in the nematode Caenorhabditis elegans increased survival21 and the expression of genes with similar function in the fruit fly Drosophila melanogaster (yolk protein genes) is negatively correlated with lifespan22. The strong support for a role for Vg expression in regulating lifespan and fecundity in honey bees, combined with conflicting results from other species, led us to ask if honey bee Vg has different functional properties than its homologs in other invertebrates. Specifically, we asked (1) if transgenic expression of honey bee Vg in fruit flies can regulate lifespan or fecundity, and (2) if over-expression of a related gene that is endogenous in flies has effects similar that of expression of honey bee Vg.\n\nAn extensive genetic toolbox allows time and tissue-specific manipulation of gene expression in D. melanogaster in ways that are not available in other organisms, and these techniques have been used to characterize the effects of many genes on fly lifespan and fecundity23. For example, manipulations of genes involved in the insulin signaling and the target-of-rapamycin pathways have been causally linked to lifespan regulation in flies (reviewed in23,24). To our knowledge, however, the effects of Vg-family genes on aging-related traits have not been investigated in flies. Unfortunately, D. melanogaster lacks a direct homolog of honeybee Vg. Instead, gene CG31150, recently annotated as crossveinless d (cv-d), encodes a Vg-like protein that is expressed mainly in the fat body25. The coding sequence for this gene is the most similar among all Drosophila genes to Vg-encoding genes in honey bee (37% similarity), C. elegans (40% similarity), chicken (38% similarity) and zebra fish (38% similar)26. CG31150 resembles other Vg family genes in having an N-terminal Vitellogenin N domain, DUF1943 Pfam motifs, and a partial von Willebrand Factor D (VWD) domain near the C terminus25. The biological functions of CG31150 are largely unknown, but it was recently implicated in lipid transport and bone morphogenic protein (BMP) signaling.\n\nWe used the bipartite GAL4/UAS system to manipulate the expression of both honey bee Vg and the endogenous CG31150 gene in flies, and to assess the effects of these manipulations on fly lifespan and fecundity. In this system, expression of a transgene is under the control of a promoter region derived from yeast, the Upstream Activation Sequence (UAS). This promoter region activates transgene expression only when it is bound to the GAL4 protein. Tissue-specific promoters enable spatial control of GAL4 protein production27. Temporal control of transgene expression can be achieved in several ways28, but one of the most convenient methods is using constructs in which the GAL4 DNA binding domain is fused to a progesterone receptor transcriptional activation domain, and is therefore only activated by systemic application of a progesterone-receptor ligand29–31. We used RU486 (mifepristone)-induced GAL4 drivers (also called the GeneSwitch system) to manipulate our target genes in adult fat body32. A major advantage of this system is that control flies (not exposed to RU486) have exactly the same genotype as transgene-expressing flies, so genetic background differences cannot contribute to differences in phenotype33. We manipulated the expression of both the honey bee Vg and D. melanogaster CG31150 in order to determine whether either or both can regulate lifespan or fecundity in flies. We included two constructs of each gene to account for potential position effects. We also included a series of controls to take into account possible phenotypic effects of RU486 or of the expression of the GAL4 protein34,35.\n\n\nMaterials and methods\n\nThe GeneSwitch driver strain S106 (w1118; P{Switch1}106) and UAS-GFP strain (P{UAS-GFP.VALIUM10}attP2) were obtained from Bloomington Drosophila Stock Center. To control for position effects, we used two different transgenic strains for both Vg and CG31150, with the transgene inserted onto different chromosomes: CG31150-2 (w1118, UAS-CG31150-2/FM6), CG31150-4 (w1118, UAS-CG31150-4/TM3), Vg-1 (w1118, UAS-Apis VGD13-1/FM6), Vg-2 (w1118, UAS-Apis VGD13-2/TM3). These constructs were created by Eric Spana at the Duke University Model Systems Genomics Core Facility as follows. For the Vg constructs, four cDNAs were identified as Apis Vg from database searches, and obtained from RIKEN. All four were sequenced on both strands, and one, BH10008D13 was chosen for subsequent cloning as it matched the published sequence best. Site directed mutagenesis was used to remove an ATG from the parental cloning vector (CATTATACGAAGTTAGGGATCAGGCCAAATCGGCCG where the underlined G marks the nucleotide that was changed from a T) so that a NotI/KpnI double digest would excise the Vg cDNA from the vector and not contain an incorrect start codon. The NotI/KpnI fragment was then ligated into a NotI/KpnI linearized pUAST. The subsequent clone was verified by sequencing on both strands, and pUAST-Apis_Vg was then transformed into Drosophila as described below. For the CG31150 constructs, cDNA GH05619 was available in the “Gold Clone” collection and was obtained from Robin Wharton (Duke University Medical Center). The cDNA was excised from GH05619 using EcoRV & XhoI and ligated into pUAST that had been digested by EcoRI, filled in with Klenow to make a blunt end, and then digested with XhoI to make compatible ends with the insert. The pUAST-CG31150 plasmid insert was sequenced on both strands. pUAST-Apis Vg and pUAST-CG31150 plasmid inserts were transformed into w1118 by Model System Genomics of Duke University by standard techniques. All transgenic flies were mapped and balanced. Transgene sequences are shown in Supplementary Table S1.\n\nOverexpression of the target gene in each of the four genotypes was achieved by crossing each transgenic stock with the GeneSwitch strain S106. Strain S106 drives expression specifically in the fat body and also in the digestive system32. This driver has been widely used in aging studies; for example, over-expression of dfoxo and dilp6 using this driver have been shown to increase lifespan36,37. To produce female flies that expressed the target transgene, virgin females from the S106 stock were crossed to males from each of the UAS-transgenic strains. Half of the female offspring from this cross were reared with RU486-supplemented media throughout adult life (see below), thus inducing transgene expression, but only during the adult stage. The remaining females from each cross were reared on media supplemented only with RU486 vehicle (ethanol), and should not have expressed the transgene. Therefore each group of flies that expressed one of the target transgenes had a genotype-matched control that differed only with respect to whether or not it was fed RU486 or vehicle.\n\nTo determine if RU486 exposure (or the GAL4 expression it induced) caused any change in lifespan or fecundity in these flies, we used an additional set of controls. We crossed the S106 driver strain to UAS-GFP flies, and exposed half of the female offspring to RU486 and half only to the ethanol vehicle. GFP is widely used as a reporter in D. melanogaster, and is believed to be non-toxic and not to influence endogenous gene expression at any stage during fly development38. Flies were kept on a 12:12 light: dark cycle at 25°C for their entire lifespan.\n\nAn RU486 (Mifepristone, Sigma, St. Louis, MO, USA) stock solution of 25mg/ml was made in 100% ethanol. Appropriate volumes of the stock were diluted with water to reach a final concentration of 65µg/ml, and 300µl of the diluted solution was added onto the surface of standard fly food (1.6% yeast, 0.92% soy flour, 6.76% cornmeal, 4.28% malt, 0.61% agar, 0.25% tegosept, 7.12% corn syrup, 0.61% propionic acid and 77.85% water).\n\nSimilar drug concentration and delivery methods have been widely used in aging studies32,35,36, and our pilot data indicated that this concentration produced robust expression levels in the range that we desired (5–15 fold increase over non-induced controls). For the control food, an equal amount of 100% ethanol was diluted with water and added to the surface of standard fly food. Each type of food was made fresh twice per week. The vials were allowed to air dry for 48 hours prior to each transfer.\n\nFor each genotype tested, female offspring were collected within 24 hours of emergence and split equally into experimental and control groups. For each genotype, 96 females were reared on media supplemented with RU486, and 96 were reared on control media. Flies were housed in standard rearing vials (VWR) at a density of 6 females per vial. All rearing vials were placed randomly in 10 × 10 vial racks to control for possible position effects. The flies were transferred to new media twice per week, and were counted nearly every day. In addition, 3 replicates of 10 females each from each treatment group per genotype were collected separately to be assayed for target gene expression. For these replicates, flies were sampled at 7 days post eclosion; GFP/S106 flies were examined for GFP expression using Zeiss LSM 5 PASCA fluorescent microscopy (Zeiss, Germany), and Vg/S106 and CG31150/S106 flies were flash frozen on dry ice for quantitative PCR assays.\n\nFemales described above were placed in vials with males from a wild-type laboratory strain (described Remolina et al.39) when the females were 1 day old. 6 virgin females and 4 males were placed into each vial. The males were removed after 3 days, and females were transferred to new vials. Flies were transferred twice per week. After each transfer, the old vials were kept at 25°C, and all offspring eclosing within 14 days after the adults were removed and counted.\n\nFor each genotype, we pooled 10 flies from each replicate and homogenized whole flies using cordless motors (VWR) and RNase-free pellet pestles (Kimbel Chase). Total RNA was extracted from whole bodies using a PicoPure RNA isolation kit (Arcturus), according to the manufacturer's recommended protocol. RNA purity and quantity were measured using a Nanodrop spectrophotometer (Thermo Scientific). RNA was DNase-treated with a DNA-free RNA kit (Zymo Research) and reverse transcribed using the Superscript III (Invitrogen). We conducted qRT-PCR using SYBR green master mix (Applied Biosystems), and ribosomal gene Rp49 as an endogenous control. The primer sequences were:\n\nRp49: F ‘TCGAACCAGGCGGGCATATTGT’, R ‘TCGAACCAGGCGGGCATATTGT’;\n\nVg: F ‘AGCTGGTCGGGGCTACGTCC’, R ‘TAAGGGCGTCGGAGGGGACC’;\n\nCG31150: F ‘ACGGACACCGACTTCTGTCCCA’, R ‘TCGAACCAGGCGGGCATATTGT’\n\nFor transgenic GFP flies (UAS-GFP/S106), expression of GFP was assessed by fluorescent microscopy.\n\nFor each transgenic line, we compared Kaplan–Meier (product-limit) survival estimates between RU486- and vehicle-fed flies by using log-rank and generalized Wilcoxon chi-square tests of the homogeneity of survival functions between groups. These tests were conducted using JMP Pro 10 statistical software (SAS Institute Inc., Cary, North Carolina, United States). Log-rank tests are more sensitive to survival differences that occur late in life while Wilcoxon tests are more sensitive to differences earlier in life40. Results of both tests were consistent in all the analyses reported here, so we report only the log-rank test results. To determine if activation of transgene expression by RU486 had different effects in different genotypes, we use used Cox proportional hazard models with predictor variables that included genotype, RU486 treatment, and genotype-by-treatment interaction. A significant interaction in this model would indicate that genotypes responded differently to RU486 treatment. These tests were conducted using the phreg procedure of SAS v. 9.3. Flies that died accidentally or escaped during transfers were recorded as censored on the day they died or escaped. For fecundity assays, the mean age-specific fecundity for each vial was calculated by dividing the number of offspring by the number of female parents alive in the vial. We then compared fecundity among genotypes and treatment groups using repeated-measures ANOVA of mean fecundity, with age of the females as the repeated measure. These analyses were conducted in JMP Pro 10.\n\n\nResults\n\nFlies carrying S106 GAL4 driver and each of the target genes, Vg-1, Vg-2, CG31150-2, CG31150-4, that were fed on RU486 had a >5 fold increase in target mRNA expression compared with their genetically matched controls (Table 1). In addition, all 30 GFP/S106 flies that were fed on RU486 showed fluorescent green under the fluorescent microscope while none of their genetically matched controls did (data not shown).\n\n1Target gene mRNA expression level is derived by relative quantification (RQ) after normalizing to Rp49.\n\n2Fold change in target gene expression between RU486-exposed flies and the genetically matched non-fed control.\n\nTo test whether Vg or CG31150 overexpression affected lifespan in Drosophila, we examined survival of flies with overexpression of Apis Vg and Drosophila CG31150. Contradictory to the hypothesis that an increase in the target gene expression would increase lifespan, we observed no significant differences between control and overexpression flies for any of the four experimental genotypes tested (Table 2, Figure 1a–1d). A proportional-hazards model indicated that there were no significant differences among the four Vg and CG31150 genotypes in their response to transgene activation (χ2 = 0.97, df=3, p=0.81 for genotype-by-RU486 treatment interaction). As an additional control, we assessed whether the RU486 treatment itself was associated with changes in lifespan by comparing RU486-fed and vehicle-fed UAS-GFP/S106 flies. Surprisingly, the RU486-fed flies had a significantly longer lifespan than the controls in this comparison, indicating that RU486 itself (or the GAL4 expression induced by it) had a positive effect on lifespan in this assay (Table 2, Figure 1e). Thus, the lack of lifespan-extending effects of Vg and CG31150 transgene expression were not due to confounding effects of the method of induction because this method appeared to increase, rather than decrease lifespan in this experiment.\n\nNo significant lifespan difference was observed between Vg or CG31150 overexpressed flies and their genetically matched control (a–d). The negative control GFP/S106 had significantly longer lifespan when fed with RU486 (e).\n\nTo determine if transgene expression had any effect on lifespan in the Vg and CG31150 lines, after accounting for the lifespan-extending effects of RU486 exposure, we compared the effects of RU486 across all genotypes. Proportional hazards models indicated that there was significant heterogeneity among genotypes in their response to RU486 treatment overall (χ2 = 11.2, df=4, p=0.02), with RU486 exposure associated with significantly higher mortality in the Vg and CG3115 transgenic flies than in the GFP transgenic flies in each pairwise comparison (χ2 = 6.5, df=1, p=0.01, hazard ratio = 1.77 for CG31150-2 vs GFP; χ2 = 7.8, df=1, p=0.005, hazard ratio = 1.85 for CG31150-4 vs GFP; χ2 = 8.23, df=1, p<0.005, hazard ratio = 1.87 for Vg-1 vs GFP; χ2 = 4.0, df=1, p<0.05, hazard ratio = 1.56 for Vg-2 vs GFP). These results suggest that Vg and CG31550 transgene over-expression decrease fly lifespan because (1) GFP over-expression is unlikely to increase lifespan, and in one study was shown to decrease lifespan41, and (2) if transgene expression had no effect in Vg and CG31550 lines, we should have observed a similar increase in lifespan in the RU486-fed flies driven solely by lifespan-extending effects of RU486 and/or GAL4 expression.\n\nTo investigate whether overexpression of Vg or CG31150 increased fecundity in Drosophila, we compared age-specific and lifetime fecundity between flies overexpressing these genes and genetically matched controls. We observed a significant overall reduction in fecundity of the RU486-fed flies in all UAS/S106 genotypes, including the control genotype GFP/S106 (Table 3, Figure 2). The difference between treatments was greatest during mid-life at ages of peak egg-laying, and less at early and late ages (significant age-by-treatment effects were observed in every genotype, Table 3, Figure 2).\n\nFor all of the 5 genotypes tested, the overexpression flies had lower lifetime fecundity than the control. Age × treatment effects were also significant for all phenotypes.\n\nIn two-way ANOVA models that included all genotypes, there was no significant genotype-by-treatment interaction (F=0.111, df=4,149, p=0.978) indicating that all transgenic lines, including GFP/S106, responded similarly to RU486 exposure. It therefore appears likely that the decline in fecundity in transgene-expressing flies is caused by RU486 itself or by the expression of GAL4, rather than target-gene overexpression.\n\n\nDiscussion\n\nIn this experiment, Vg and CG31150 overexpressing flies and their genetically matched controls did not differ significantly in lifespan. However, our negative control GFP/S106 showed significant increase in lifespan when fed RU486. The extended lifespan in the GFP/S106 flies is likely due to the supplement of RU486 in the diet or expression of the GAL4 protein, rather than GFP overexpression. If this lifespan-extending effect of RU486 and/or GAL4 activation also occurred in Vg- and CG31150-overexpresing lines we investigated, then over-expression of these genes likely had a corresponding negative effect on lifespan.\n\nVg and CG31150 overexpressing flies showed reduced fecundity comparing with their genetically matched controls. GFP/S106 flies fed with RU486 showed a quantitatively similar reduction in fecundity. These results suggest that over-expression of Vg and CG31150 do not affect fecundity in Drosophila. Instead the effect is attributable to RU486, the expression of GAL4, or both.\n\nTarone et al. (2012)42 reported that expression levels of Drosophila yp family genes were negatively correlated with longevity among genotypes derived from a natural D. melanogaster population. These genes, while unrelated to Vitellogenins, are thought to carry out some functions normally associated with Vitellogenins (e.g., they are believed to comprise the major storage protein in fly embryos)26,43. Similarly, our results suggest that Vg-family gene expression affects fly lifespan. This result contrasts with effects of Vg expression in hymenoptera15,16,19,44,45. The specific pathway through which these genes affect fly lifespan is not known, but our results suggest that it was unlikely due to trade-offs with fecundity, since no increase in fecundity was detected in transgene-overexpressing flies.\n\nIn addition to CG31150, three other Vg-family genes have been described in D. melanogaster26, but their effects on aging have not been characterized. These genes include the major hemolymph lipid carrier (apolpp) and regulators of interorgan lipid transport (mtp and apoLTP). Products of these genes might therefore play major roles in regulating energy allocation and life history patterns, and future studies of their effects on aging-related traits are warranted.\n\nIn addition to testing effects of Vg and CG31150, our results highlighted the importance of using multiple controls in aging studies. If we had not used the GFP/S106 controls, we would not have identified the effects of RU486 exposure in increasing female fly lifespan and decreasing fecundity, and we would have concluded that expression of our target genes had no effect on lifespan and that they decreased fecundity. Aging-related phenotypes are sensitive to many environmental variables, genetic background, and interaction between genotype and environment. A recent meta-analysis of Drosophila aging experiments reported that transgenic constructs of the GAL4/UAS expression system that should have had no phenotypic effects (GAL4 alone, UAS alone, or noninduced GeneSwitch constructs) significantly extended lifespan in the w1118 genetic background46, consistent with our results. Other studies have reported significant reductions in lifespan caused by RU486 exposure in some genotypes35, suggesting that strain-specific effects of RU486 exposure are common. Experiments using inducible transgenic constructs therefore require multiple sets of controls to reliably assay genetic regulation of aging phenotypes.\n\n\nData availability\n\nF1000Research: Dataset 1. Drosophila lifespan and fecundity data set, 10.5256/f1000research.3975.d2831147", "appendix": "Author contributions\n\n\n\nY. Ren and K. A. Hughes designed the experiments, analyzed the data and wrote the manuscript. Y. Ren performed all experiments. Both authors agreed with the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by National Science Foundation grant DEB 0848337 and National Institutes of Health grant AG022824 to KAH.\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 G. Robinson for providing Apis Vg clones and for helpful discussion in planning this research, B. Washburn for qRT-PCR help, M. Arbeitman for providing UAS/GFP stock, and B. Thedy, M. Zornes and K. Post for help with fly husbandry.\n\n\nReferences\n\nFinch C: Longevity, Senescence, and the Genome. Chicago: University of Chicago. 1990. Reference Source\n\nRose MR: Evolutionary biology of aging. New York: Oxford UP, 1991. Reference Source\n\nTamura T, Chiang A, Ito N, et al.: “Aging specifically impairs amnesiac-dependent memory in Drosophila”. Neuron. 2003; 40(5): 1003–11. PubMed Abstract \| Publisher Full Text\n\nZerofsky M, Harel E, Silverman N, et al.: “Aging of the innate immune response in Drosophila melanogaster”. Aging Cell. 2005; 4(2): 103–08. PubMed Abstract \| Publisher Full Text\n\nFang M, Roscoe F, Sigal LJ: “Age-dependent susceptibility to a viral disease due to decreased natural killer cell numbers and trafficking”. J Exp Med. 2010; 207(11): 2369–381. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHughes KA: “Mutation and the evolution of ageing: from biometrics to system genetics”. Philos Trans R Soc Lond B Biol Sci. 2010; 365(1544): 1273–79. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGems D, Partridge L: “Genetics of longevity in model organisms: debates and paradigm shifts”. Annu Rev Physiol. 2013; 75(1): 621–44. PubMed Abstract \| Publisher Full Text\n\nStearns SC: The Evolution of Life Histories. Oxford: Oxford UP, 1992. Reference Source\n\nZera AJ, Harshman LG: “The Physiology of Life History Trade-offs in Animals”. Annu Rev Ecol Syst. 2001; 32(1): 95–126. Publisher Full Text\n\nRoff DA: The Evolution of Life Histories: Theory and Analysis. New York: Chapman & Hall, 1992. Reference Source\n\nKeller L, Jemielity S: “Social insects as a model to study the molecular basis of ageing”. Exp Gerontol. 2006; 41(6): 553–56. PubMed Abstract \| Publisher Full Text\n\nKeller L, Passera L: “Fecundity of Ant Queens in Relation to Their Age and the Mode of Colony Founding”. Insectes Sociaux. 1990; 37(2): 116–30. Publisher Full Text\n\nPage RE Jr, Peng YS: “Aging and development in social insects with emphasis on the honey bee, Apis mellifera L”. Exp Gerontol. 2001; 36(4–6): 695–711. PubMed Abstract \| Publisher Full Text\n\nKeller L, Genoud M: “Extraordinary Lifespans in Ants: a Test of Evolutionary Theories of Ageing”. Nature. 1997; 389: 958–960. Publisher Full Text\n\nCorona M, Velarde RA, Remolina S, et al.: “Vitellogenin, juvenile hormone, insulin signaling, and queen honey bee longevity”. Proc Natl Acad Sci U S A. 2007; 104(17): 7128–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmdam GV, Hartfelder K, Norberg K, et al.: “Altered physiology in worker honey bees (Hymenoptera: Apidae) infected with the mite Varroa destructor (Acari: Varroidae); a factor in colony loss during overwintering?” J Econ Entomol. 2004; 97(3): 741–747. PubMed Abstract \| Publisher Full Text\n\nMunch D, Amdam GV: “The curious case of aging plasticity in honey bees”. FEBS Lett. 2010; 584(12): 2496–2503. PubMed Abstract \| Publisher Full Text\n\nMunch D, Kreibich CD, Amdam GV: “Aging and Its modulation in a long-lived worker caste of the honey bee”. J Exp Biol. 2013; 216(Pt 9): 1638–649. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeehuus SC, Krekling T, Amdam GV: “Cellular senescence in honey bee brain is largely independent of chronological age”. Exp Gerontol. 2006; 41(11): 1117–125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelson CM, Ihle KE, Fondrk MK, et al.: “The gene vitellogenin has multiple coordinating effects on social organization”. PLoS Biol. 2007; 5(3): e62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMurphy CT, Mccarroll SA, Bargmann CI, et al.: “Genes that act downstream of DAF-16 to influence the lifespan of Caenorhabditis elegans”. Nature. 2003; 424(6946): 277–83. PubMed Abstract \| Publisher Full Text\n\nTarone AM, Mcintyre LM, Harshman LG, et al.: “Genetic variation in the Yolk protein expression network of Drosophila melanogaster: sex-biased negative correlations with longevity”. Heredity (Edinb). 2012; 109(4): 226–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPaaby AB, Schmidt PS: “Dissecting the genetics of longevity in Drosophila melanogaster”. Fly (Austin). 2009; 3(1): 29–38. PubMed Abstract \| Publisher Full Text\n\nPartridge L, Alic N, Bjedov I, et al.: “Ageing in Drosophila: the role of the insulin/Igf and TOR signaling network”. Exp Gerontol. 2011; 46(5): 376–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen J, Honeyager SM, Schleede J, et al.: “Crossveinless d is a vitellogenin-like lipoprotein that binds BMPs and HSPGs, and is required for normal BMP signaling in the Drosophila wing”. Development. 2012; 139(12): 2170–176. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPalm W, Sampaio JL, Brankatschk M, et al.: “Lipoproteins in Drosophila melanogaster--assembly, function, and influence on tissue lipid composition”. PLoS Genet. 2012; 8(7): e1002828. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrand AH, Perrimon N: “Targeted gene expression as a means of altering cell fates and generating dominant phenotypes”. Development. 1993; 118(2): 401–15. PubMed Abstract\n\nDuffy JB: “GAL4 system in Drosophila: a fly geneticist's swiss army knife”. Genesis. 2002; 34(1–2): 1–15. PubMed Abstract \| Publisher Full Text\n\nOsterwalder T, Yoon KS, White BH, et al.: “A conditional tissue-specific transgene expression system using inducible GAL4”. Proc Natl Acad Sci U S A. 2001; 98(22): 12596–2601. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoman G, Endo K, Zong L, et al.: “P[Switch], a system for spatial and temporal control of gene expression in Drosophila melanogaster”. Proc Natl Acad Sci U S A. 2001; 98(22): 12602–607. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNicholson L, Singh GK, Osterwalder T, et al.: “Spatial and temporal control of gene expression in Drosophila using the inducible GeneSwitch GAL4 system. I. Screen for larval nervous system drivers”. Genetics. 2008; 178(1): 215–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoirier L, Shane A, Zheng J, et al.: “Characterization of the Drosophila gene-switch system in aging studies: a cautionary tale”. Aging Cell. 2008; 7(5): 758–70. PubMed Abstract \| Publisher Full Text\n\nPartridge L, Gems D: “Benchmarks for ageing studies”. Nature. 2007; 450(7167): 165–67. PubMed Abstract \| Publisher Full Text\n\nRen C, Finkel SE, Tower J: “Conditional inhibition of autophagy genes in adult Drosophila impairs immunity without compromising longevity”. Exp Gerontol. 2009; 44(3): 228–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShen J, Curtis C, Tavaré S, et al.: “A screen of apoptosis and senescence regulatory genes for life span effects when over-expressed in Drosophila”. Aging (Albany NY). 2009; 1(2): 191–211. PubMed Abstract \| Free Full Text\n\nGiannakou ME, Goss M, Jünger MA, et al.: “Long-lived Drosophila with overexpressed dFOXO in adult fat body”. Science. 2004; 305(5682): 361. PubMed Abstract \| Publisher Full Text\n\nBai H, Kang P, Tatar M: “Drosophila insulin-like peptide-6 (dilp6) expression from fat body extends lifespan and represses secretion of Drosophila insulin-like peptide-2 from the brain”. Aging Cell. 2012; 11(6): 978–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChalfie M: “Green Fluorescent Protein as a Marker for Gene Expression”. Trends Genet. 1994; 10(5): 151. Publisher Full Text\n\nRemolina SC, Chang PL, Leips J, et al.: “Genomic basis of aging and life-history evolution in Drosophila melanogaster”. Evolution. 2012; 66(11): 3390–403. PubMed Abstract \| Publisher Full Text\n\nAllison PD: Survival Analysis Using SAS System: A Practical Guide. Cary, NC: SAS Institute, 1995. Reference Source\n\nMawhinney RM, Staveley BE: “Expression of GFP can influence aging and climbing ability in Drosophila”. Genet Mol Res. 2011; 10(1): 494–505. PubMed Abstract \| Publisher Full Text\n\nTarone AM, McIntyre LM, Harshman LG, et al.: “Genetic variation in the Yolk protein expression network of Drosophila melanogaster: sex-biased negative correlations with longevity”. Heredity (Edinb). 2012; 109(4): 226–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBownes M: “Why is there sequence similarity between insect yolk proteins and vertebrate lipases?” J Lipid Res. 1992; 33(6): 777–90. PubMed Abstract\n\nLandis GN, Abdueva D, Skvortsov D, et al.: “Similar gene expression patterns characterize aging and oxidative stress in Drosophila melanogaster”. Proc Natl Acad Sci U S A. 2004; 101(20): 7663–7668. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrandt BW, Zwaan BJ, Beekman M, et al.: Shuttling between species for pathways of lifespan regulation: a central role for the Vitellogenin gene family? BioEssays. 2005; 27(3): 339–346. PubMed Abstract \| Publisher Full Text\n\nZiehm M, Piper MD, Thornton JM: Analysing variation in Drosophila aging across independent experimental studies: a meta-analysis of survival data. Aging Cell. 2013; 12(5): 917–922. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRen Y, Hughes KA: Drosophila lifespan and fecundity data set. F1000Research. 2014. Data Source" }	[ { "id": "5095", "date": "23 Jun 2014", "name": "Joel Parker", "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, interesting and well executed study showing similarities between honey bee and Drosophila vitellogenin, but with neither extending lifespan nor effecting fecundity when over expressed in flies. The study also offers an important lesson in using controls in UAS-GAL4 systems. The purpose was to ask whether the most closely related gene in Drosophila to the vitellogenin gene in honey bee is also involved in aging. This was a long shot from the level of homology but a question worth asking. The results were basically negative but the authors also found very clear and striking effects in the controls that anyone using the UAS-GAL4 system should be aware of.There are several points that the authors should consider addressing in their revision:Are all five of the transgenic sets on the same genetic background? It is not clear from the methods where the w1118 flies were sourced and there is no mention of backcrossing. If they came from different sources it is possible that the backgrounds might be different making comparisons across the five less meaningful. Even if the backgrounds are not exactly the same the on/off nature of the UAS-GAL4 system in the same lines still means the effects are real within each pair. Another interpretation of the longevity results is that the honey bee Vg and fly homolog both blocked the effect of GAL4 instead of the authors’ interpretation that over expression of both of these genes decrease lifespan. This is an important because it might be overreaching to interpret the results as a shortening of lifespan in general unless one assumes the effects of GAL4 and the Vg genes are independent and additive. The authors should start off the discussion with an important positive result they neglected to point out, that both the honey bee and the fly Vg genes showed exactly the same effects on both longevity and fecundity. The consistent results from the two Vg’s might be saying they share some functions. Hence the previously seen effect on aging is only in honey bees and not flies. I realise that this is a weak supposition given that fecundity was not changed and that only a control effect may have been reversed in the aging experiment, but the fact that both gave exactly the same results is consistent with the central hypothesis that the two have homologous functions. I think there is a mis-wording at the end of the second sentence of the 5th paragraph in the Discussion. Remove “that they decreased”. The point is that fecundity was not reduced when you consider the effect seen in the control. Early on in the honeybee vitellogenin story there was a suggestion that it increased lifespan by acting as an anti-oxidant. If this was the case then one might have expected to see an increase in lifespan in the Drosophila as well. The fact that this did not happen suggests that honeybee vitellogenin is working in bees in another manner. On the other hand, antioxidants can decrease lifespan in some Drosophila genetic backgrounds in which case they see Vg’s functioning in exactly the same way in both flies and bees (assuming the GAL4 is additive). Although this is all speculative, it still might be worth going into a bit more detail about what is known and thought about the molecular mechanisms underpinning how Vg effects aging in honey bees and more clearly define the a priori predictions. Summary:The title is appropriateAbstract is adequate Everything is well explained and presented The conclusions are sound and sufficiently conservative, I do make some suggestions above.", "responses": [] }, { "id": "5099", "date": "24 Jun 2014", "name": "John Tower", "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\nPrevious reports of a positive correlation between Vg gene expression and longevity in the hymenoptera prompted this interesting and timely test of Vg gene over-expression in Drosophila. The data show that there is no simple strong benefit to life span when over-expressing Vg genes in the Drosophila adult fat body, using the S1-106 Gene-Switch driver strain. In addition, because a life span increase was observed in a control cross involving over-expression of GFP, the results suggest possible effects of the drug, the activated Gene-Switch transcription factor, the GFP, or the variability of the assay. These results underscore the importance of careful controls for possible life span effects on the conditional gene expression system itself, independent of the specific identity of the gene being over-expressed. The data are clearly presented and support the conclusions above, however, the generality of the conclusions must be tempered somewhat given the limited scope of the study. In previous Drosophila studies life span extension was observed when transgenes were expressed with certain drivers, but not others, including some cases where the drivers tested had similar tissue specificities. Those results indicate that the precise tissue-specificity and/or level of transgene expression can be important. In the present study the Vg transgenes were expressed using a Gene-Switch driver strain that yields expression in only a subset of the fat-body tissues, so it remains possible that over-expression of the Vg genes using a driver with more extensive expression pattern in the fat body and/or other tissues might create a life span increase. The conclusions regarding the possible cause of the life span increase in the control cross are based on the assumption that over-expressed GFP has little or no phenotypic effect. However the one Drosophila study cited reports a significant (negative) effect of GFP expression on aging phenotypes (using different drivers). If GFP can have negative effects in some tissues it seems possible it might have positive effects when expressed in other tissues (for example, the “negative” effect of killing the IPCs or the germ cells can increase fly life span). For these reasons it is not clear whether the life span increase observed in the present study results from the drug, the activated Gene-Switch, or the GFP. Finally, it should be noted that this result was not replicated. These considerations limit the support for a possible negative life span effect of the Vg transgenes as suggested in the Abstract and Discussion.", "responses": [] } ]	1	https://f1000research.com/articles/3-125
https://f1000research.com/articles/3-18/v1	21 Jan 14	{ "type": "Research Article", "title": "Mutation extraction tools can be combined for robust recognition of genetic variants in the literature", "authors": [ "Antonio Jimeno Yepes", "Karin Verspoor", "Karin Verspoor" ], "abstract": "As the cost of genomic sequencing continues to fall, the amount of data being collected and studied for the purpose of understanding the genetic basis of disease is increasing dramatically. Much of the source information relevant to such efforts is available only from unstructured sources such as the scientific literature, and significant resources are expended in manually curating and structuring the information in the literature. As such, there have been a number of systems developed to target automatic extraction of mutations and other genetic variation from the literature using text mining tools. We have performed a broad survey of the existing publicly available tools for extraction of genetic variants from the scientific literature. We consider not just one tool but a number of different tools, individually and in combination, and apply the tools in two scenarios. First, they are compared in an intrinsic evaluation context, where the tools are tested for their ability to identify specific mentions of genetic variants in a corpus of manually annotated papers, the Variome corpus. Second, they are compared in an extrinsic evaluation context based on our previous study of text mining support for curation of the COSMIC and InSiGHT databases. Our results demonstrate that no single tool covers the full range of genetic variants mentioned in the literature. Rather, several tools have complementary coverage and can be used together effectively. In the intrinsic evaluation on the Variome corpus, the combined performance is above 0.93 in F-measure, while in the extrinsic evaluation the combined recall performance is above 0.71 for COSMIC and above 0.62 for InSiGHT, a substantial improvement over the performance of any individual tool. Based on the analysis of these results, we suggest several directions for the improvement of text mining tools for genetic variant extraction from the literature.", "keywords": [ "amino acids", "DNA", "deletions" ], "content": "Introduction\n\nAs the cost of genomic sequencing continues to fall, the amount of data being collected and studied for the purpose of understanding the genetic basis of disease is increasing dramatically. There are large-scale efforts to catalog the results of this research in structured databases, including in the Online Mendelian Inheritance in Man (OMIM) database1 and the Human Gene Mutation Database (HGMD)2. Much of the source information relevant to such efforts is available only from unstructured sources such as the scientific literature, and significant resources are expended in manually curating and structuring the information in the literature. As such, there have been a number of systems developed to target automatic extraction of mutations and other genetic variation from the literature using text mining tools3–9, inter alia. Such tools have been shown to perform well, benefiting from a well-defined target vocabulary (nucleic and amino acids), the availability of reference sequences for position validation, and increasing adoption of standard nomenclature such as the Human Genome Variation Society (HGVS) format10. The natural language descriptors of genetic variation are fairly consistent, and lend themselves well to automated processing.\n\nIn previous work11,12, we assessed one of these tools, the Extractor of Mutations (EMU) tool6, for its ability to identify the genetic variant information that had been manually curated in the COSMIC13 and the International Society for Gastro-intestinal Hereditary Tumours (InSiGHT)14 databases from targeted literature sources. That work found very low recall for the text mining tool when considering the narrative content of publications alone, and identified processing of the supplementary material associated with publications as a critical component of an approach to automated genetic variant curation from the literature.\n\nIn this work, we perform a broad survey of the existing publicly available tools for extraction of genetic variants from the scientific literature. We consider not just one tool but a number of different tools, individually and in combination, and apply the tools in two scenarios. First, they are compared in an intrinsic evaluation context, where the tools are tested for their ability to identify specific mentions of genetic variants in a corpus of manually annotated papers, the Variome corpus15. Unlike previous test corpora, this corpus was not designed exclusively for the purpose of testing mutation extraction tools and hence is a better test of real-world applicability than prior corpora. Second, they are compared in an extrinsic evaluation context based on our previous study with COSMIC and InSiGHT. Our results demonstrate that several of the tools have complementary coverage and can be used together effectively. This study suggests several directions for the improvement of text mining tools for genetic variant extraction from the literature.\n\n\nBackground\n\nText mining of mutations in the scientific literature has been addressed by several tools, including MutationMiner3, MarkerInfoFinder16, EMU (Extractor of Mutations)6, MutationFinder4, tmVar9, and SETH17. A summary of previous work can be found in Naderi and Witte (2012)7. These tools have been shown to achieve a performance over 0.90 in F1 measure, and in some cases almost perfect Precision/Recall, on intrinsic evaluations. There are also several corpora that are publicly available to support intrinsic evaluation of mutation extraction tools4,6,16,18–20. On the other hand, cross-comparison of these tools has been limited. The most commonly used is the corpus provided with the Mutation Finder tool, which covers protein variants. Some of these tools have also been used to reproduce the information curated in existing databases about genetic variants, allowing for extrinsic evaluation of the mutation extraction tools.\n\nIn the following sections, we present the tools for genetic variant extraction that we have considered in our study. Each is a publicly available tool. The tools are introduced and their published results in intrinsic evaluation are presented. Results are presented in terms of precision (P), recall (R) and F1 measure (F).\n\nMutationFinder (MF) was one of the earliest tools developed for extraction of mutations. This tool performs point mutation extraction based on a set of regular expressions4. The coverage of this tool is thus limited compared to the other ones. A corpus, the MutationFinder corpus, was established to guide the construction of the patterns. The development data set is made up of 605 point mutation mentions in 305 abstracts selected randomly from primary citations in PDB. The evaluation data set is made up of 910 point mutation mentions in 508 abstracts annotated by two of the authors, not involved in the development of the system. Mean pairwise interannotator agreement, calculated on the fifty overlapping abstracts, was 94%4. Performance of MF in the extracted mutations is P: 0.984 R: 0.817 F: 0.893.\n\nOpen Mutation Miner (OMM) is a tool7 that extracts protein mutation mentions and maps them to their properties in the OMM impact ontology, which covers protein mutation types (insertion, deletion, point mutation), protein types and the impact of the mutation. The extraction of mutations component couples grammar rules with a normalization step, e.g. transformation into single-letter format. The system is combined with MutationFinder to identify variants in natural language. OMM has been developed under the GATE platform and it is available as a JAPE-based mutation tagging component. Mutation extraction was evaluated on a set of 11 full text articles, with an average performance of P:0.99, R:0.96 and F:0.977.\n\nThe Extractor of Mutations (EMU) tool6 was designed to capture a broader range of mutations than other tools available when it was developed and hence is a better fit for the variants we might expect to find. It identifies protein and DNA point mutations, Single Nucleotide Polymorphism Database (dbSNP) identifiers21 (RSIDs), and DNA insertions and deletions. In addition, it links the mutations to the proteins and genes that appear in text and performs sequence verifications using existing sequence databases to increase the precision of the annotations. EMU has been shown to have a performance of 0.92 F1 measure on an intrinsic evaluation6, i.e., it has high recall and high accuracy.\n\ntmVar9 (http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/pub/tmVar) is a recently released mutation extraction tool based on a conditional random field model used with a special set of features, which has shown to perform better than MutationFinder on the MutationFinder corpus. tmVar has been trained on 334 newly annotated citations, different from the MutationFinder corpus, and evaluated using the MutationFinder test set (P: 0.9880, R: 0.8962, F: 0.9398) and using their own data set made up of 134 manually annotated test citations (P: 0.9138, R: 0.9140, F: 0.9139)9.\n\nSNP Extraction Tool for Human Variations (SETH) (http://rockt.github.io/SETH) implements an Extended Backus–Naur Form (EBNF) grammar proposed by Laros et al.22 to identify mentions of mutation that obey the HGVS nomenclature. Since mentions in text might not follow the HGVS nomenclature, SETH integrates MutationFinder to extend its coverage. SETH returns whether a mutation is a DNA or protein variant and the type of variant (e.g. deletion).\n\nSETH has been evaluated on several corpora, as reported on the SETH web site. Among other corpora, SETH evaluation has been performed on the MutationFinder test set (Precision 0.97 Recall 0.83 F: 0.89), the tmVar corpus (P: 0.94, R: 0.81, F: 0.87), the Thomas et al. corpus23 (P: 0.95, R: 0.58, F: 0.72) and the Osiris corpus24 (P: 0.98, R: 0.85, F: 0.91).\n\nIntrinsic evaluation: annotated corpora. There are several text corpora that have been made available for the evaluation of mutation extraction tools. There are corpora that focus on protein mutations alone, on protein and DNA mutations or on normalizing the mentions to dbSNP identifiers.\n\nSeveral corpora are available for the evaluation of protein mutation extraction tools. As presented above, the developers of MutationFinder4 made available a data set (http://mutationfinder.sourceforge.net) of 305 abstracts annotated with point mutations that was used for system development and 508 abstracts available for evaluation. The developers of OMM7 (http://www.semanticsoftware.info/open-mutation-miner) performed experiments on 11 full text articles annotated manually with protein mutations, although these documents are not publicly available for distribution, the manual annotations are available with the tool download. Other corpora exist annotated with protein residue information either manually annotated18,19 or prepared using automatic methods20.\n\nThere are corpora available that contain both protein and DNA mutations. The EMU corpus6 (http://bioinf.umbc.edu/EMU/ftp) was developed for annotation of mutations related to prostate cancer. This data set was developed by querying MEDLINE for the medical subject heading (MeSH) Mutation and selecting citations relevant for prostate cancer based on MetaMap annotation. It contains 500 manually annotated abstracts with 95 mutations in 55 abstracts. The tmVar9 (http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/pub/tmVar) system also comes with its own annotated set. The set comprises 500 abstracts manually annotated from which 334 were used for training tmVar while the remaining 166 were used for testing it. These citations were recovered from PubMed selecting English abtracts containing novel human mutations, targeting formulaic mentions. The Variome corpus15 (http://www.opennicta.com/home/health/variome) is an annotated biomedical textual resource pertaining to human genetic variation and its relation to diseases and other entity types. At present, the corpus comprises ten double annotated full text journal publications on inherited colorectal cancer, which were selected on the basis of their relevance to the genetics of the Lynch Syndrome to support the curation of the InSiGHT database.\n\nThere are two corpora to evaluate the normalization of extracted mutations to dbSNP identifiers. OSIRIS24 (https://sites.google.com/site/laurafurlongweb/databases-and-tools/corpora) was collected from MEDLINE covering English abstracts for human mutation, limited to 2004 and 2005 and focused on specific project criteria. The annotation is performed focused on evaluation of entity recognition and of disambiguation of variation entities to dbSNP identifiers. The final version of the corpus contains 105 abstracts available with 109 normalized variants and 155 unnormalized ones. The second corpus, developed by Thomas et al.23 (http://www.scai.fraunhofer.de/snp-normalization-corpus.html), consists of 296 MEDLINE abstracts annotated with 527 mutationdbSNP id pairs. The corpus was annotated initially with MutationFinder and then manually annotated for completion and normalized to dbSNP. Mutations without a valid dbSNP identifer were removed. Only the annotations are available, while the abstracts can be downloaded from PubMed using NCBI’s E-utilities (http://www.ncbi.nlm.nih.gov/books/NBK25500).\n\nTable 1 summarizes the results of the tools presented in the previous section as reported on available corpora. We can see that not many tools have been evaluated with the same corpus, other than the MutationFinder test set. The different tools evaluated on the MutationFinder corpus show that the performace in precision is generally very high across the tools with some differences in recall. However, the MutationFinder corpus covers only a limited set of protein mutations. In this work, we perform an intrinsic evaluation of the tools using the Variome corpus as the common reference set, producing a broader comparison of the different mutation extraction tools. We will introduce this corpus in detail in the Methods section.\n\nThe results in precision, recall and F1 measure are obtained from already published studies or the tool website like SETH. A dash(-) is used to indicate that no results are available. The tools are MutationFinder (MF), OpenMutationMiner (OMM), Extractor of Mutations (EMU), tmVar and SNP Extraction Tool for Human Variations (SETH). The corpora used for evaluation are MutationFinder corpora (MF), OpenMutationMiner corpora (OMM), EMU Prostate Cancer set (PCa), the tmVar corpora, the OSIRIS corpora and the corpora made available by Thomas et al.23 (Thomas).\n\nExtrinsic evaluation: curated databases. In addition to intrinsic evaluation, there have been several efforts in trying to reproduce the information curated in mutation databases through information extraction from the literature. A broader summary of extrinsic evaluation is available in11.\n\nThere have been several efforts with varying success in recovering the curated information from variants databases. Krallinger et al.5 extracteded mutations from literature for the kinase domain from abstracts and full text showing different levels of coverage of KinMutBase25, the Swissprot Variant database26, SAAPdb27 and the COSMIC database28, for which only 6% of the mutations were recovered. Schenck et al.29 worked on a small set of articles curated in COSMIC that they annotated, recovering up to 30% of their annotated mutations. Caporaso et al.30, Nagel et al.18 and Verspoor et al.31 tried to recover information about protein mutations and residues annotated in PDB protein records with limited coverage when using abstracts and with larger, but still very limited, coverage when using full text. On the pharmacogenomics side, Rance et al.32 and Hakenberg et al.8 tried recovering variants and related drugs to reproduce the data in PharmGKB with different coverage depending on the target genes.\n\nIn a previous study11, we evaluated the ability of a mutation extraction tool to recover the curated mutations in the COSMIC and InSiGHT databases using the articles that were curated in these databases. We found, as in previous studies, that the recall considering the mutations extracted from the abstracts was very low. When considering the full text of these articles the recall increased but was still low. We found that many of the missing mutations were extracted by EMU from tables and supplementary material. In our current work, we have expanded this study by performing an intrinsic evaluation of several mutation extraction tools using the Variome corpus and performing a coverage evaluation of the tools when recovering the curated mutations from COSMIC and InSiGHT.\n\n\nMethods\n\nWe have performed the evaluation and comparison of mutation extraction tools intrinsically, using the Variome corpus15, developed in collaboration with the Human Variome Project (http://www.humanvariomeproject.org), as a reference set. For the extrinsic study, we required a curated database that includes mutations and specific links to the literature (with PubMed identifiers [PMIDs] included for each mutation). We selected the COSMIC and InSiGHT databases for our investigation. These databases are used as reference sets; the information extracted from the corresponding scientific literature is compared directly to the information curated from those articles in the databases. We normalize extracted mutation mentions to Human Genome Variation Society (HGVS) format10.\n\nThe Variome corpus15 (http://www.opennicta.com/home/health/variome) is an annotated resource of biomedical texts pertaining to human genetic variation and its relation to diseases and other related entity types. At present, the corpus comprises ten double annotated full text journal publications on inherited colorectal cancer, which were selected on the basis of their relevance to the genetics of the Lynch Syndrome to support the curation of the InSiGHT database. The annotation schema covers thirteen relations, such as gene-has-mutation, mutation-has-size and disease-related-to-bodypart; and eleven entity types, such as genomic categories (e.g., gene, mutation), phenotypic categories (e.g., disease, body-part), categories related to the occurrence of mutations in a disease (e.g., age, ethnicity), and a characteristic category for the eventual addition of relevant information. Compared to other variation corpora, the Variome corpus not only annotates mutation mentions but also other entity types, providing a larger set of relevant entities. In addition, it contains annotations for relations between the entities, which provide a more exhaustive context for the training and evaluation of text mining tools supporting the curation of genetic variant databases.\n\nThe mutation entity type captures mentions of mutations which specify changes in the protein or DNA sequence as well as mutation terms which refer to general properties of a mutation (e.g. somatic mutation) or terms specifying a mutated gene (e.g. APC+). Current mutation extraction tools are only concerned with the first type, thus extracting mentions of protein or DNA changes. We have manually catalogued the annotated mutations and identified 118 mutation instances that are annotated in the corpus. From this set, 52 are DNA mutations and 66 are protein mutations.\n\nIn this work, we expand our previous study11 and retain the original data sets for ease of comparison.\n\nCOSMIC13 (http://www.sanger.ac.uk/cosmic) contains comprehensive, curated, information on somatic mutations in human cancer. We used version v62 (from 29th of November 2012) available from COSMIC’s FTP site (ftp://ftp.sanger.ac.uk/pub/CGP/cosmic/), including mutation information curated from 9,950 unique PubMed articles, as well as Cancer Genome Project (CGP) (http://www.sanger.ac.uk/genetics/CGP) studies and international system screens (e.g. International Cancer Genome Consortium (ICGC) (http://dcc.icgc.org/web)). We identified 7,898 publications associated to mutation information in this resource. cDNA and protein mutation information is already available in HGVS format. Genes are referenced by name and by HGNC (HUGO Gene Nomenclature Committee) identifier.\n\nInSiGHT14 maintains a database of genetic variants for both Lynch Syndrome and Familial Adenomatous Polyposis. The current database only has curated mutations for four genes: MLH1, MSH2, MSH6 and PMS2. The original database was established in the 1990s with mutations reported by individual laboratories. Reports manually extracted from published literature currently comprise the majority of entries in the InSiGHT database (approximately 75%, according to the database curator), with the balance direct submissions from clinics.\n\nWe accessed the InSiGHT database on 02 January 2013 to establish our data set. The data includes variants with curated associations linked to 809 PubMed citations. The database contains information about the variants in the fields Variant/DNA and Variant/Protein. The amino acids in protein variants have been normalized to single letter amino acid abbreviation form.\n\nThere are 41 articles that have been curated both in COSMIC and InSiGHT databases. Unfortunately, none of the mutations in the overlapping articles has been curated by both databases because COSMIC is focused on somatic mutations, while InSiGHT is focused on germline mutations in only four genes.\n\nAn abstract for each PMID was retrieved from MEDLINE using NCBI’s E-utilities. Abstracts were downloaded in XML format and XML escaped characters were converted to their text characters (e.g., A–>T becomes A–>T). In the case of the COSMIC database, 17 articles did not seem to be available when querying PubMed.\n\nA small portion of PubMed is available as full text articles through the Open Access collection in PubMed Central (PMC-OA). From the 9,950 PMIDs available from the COSMIC set only 563 were available from PMC-OA. From the 809 citations for InSiGHT only 12 were available through the full text PMC-OA. This represents less than 10% of the overall set referenced by both databases.\n\nIn addition to narrative text, we have used the mutation extration tools with further content linked to the papers, which includes the tables and supplementary material and is representative of the broader full text literature33. We collected articles from COSMIC and InSiGHT that are available in the open access part of PubMed Central (PMC), since it already contains the tables in the XML of the article and there are pointers to the supplementary material. For the set of 13 articles in the InSiGHT database that could be found in PMC, InSiGHT contains 252 mutation triples. COSMIC associates 33,814 mutation triples to the 563 articles in PMC.\n\nWe extracted the tables and table captions from the full text PMC articles. The COSMIC database references 394 PMC articles with tables; 197 of these were identified as having mutations in the tables. From the InSiGHT database there are only 8 articles with tables, of which 4 contain mutations. In these articles, no mutations were found in the abstract or full text.\n\nSupplementary material was also identified from links within the PMC articles and downloaded. The InSiGHT set contains a limited number of supplementary material files (in 1/12 articles), while COSMIC has a larger number linked to the papers (in 138/563 articles). In contrast to PMC articles, available in XML following a consistent DTD (Document Type Definition), supplementary material appears in a variety of file formats. The most frequent types of supplementary material in this corpus, shown in Table 2, are, in order of frequency: MS Word documents, MS Excel spreadsheet, PDF documents, TIFF images and MS Powerpoint documents. Text from the supplementary material was extracted with Apache Tika 1.3 (http://tika.apache.org/1.3). No image processing was performed.\n\nEach row denotes one of file type. For each type, for COSMIC the number of files and the number of articles, denoted by PMIDs, is shown. In the case of the InSiGHT database, there is only one supplementary file in MS Word format linked to one article.\n\nDuring the extraction of tables and supplementary material, we realized that some PMC articles do not contain the full text in XML format but a link to a PDF version of it. From the InSiGHT collection, 4 papers out of the 13 contained only the abstract with a link to the full text in PDF format. In the COSMIC collection, the proportion was 76 papers out of 563. The PDF version for these papers has been downloaded from the European PMC mirror (http://europepmc.org), which offers a straightforward link to download the PDF files.\n\nWe have considered several state of the art mutation extraction tools in this study, as introduced in the Text mining tools for genetic variant extraction section above. We normalized the mutation mentions identified by these tools to follow the HGVS nomenclature, to be comparable to the information in the COSMIC and InSiGHT databases. This normalization required considering the specificities of each tool, thus a normalization program was prepared for each tool. Missense mutations, mutations in the DNA that result in a protein change, are normalized to amino acid (wild type), position, amino acid (mutated), using single letter amino acid abbreviations. Thus, a mutation identified with wild type amino acid Ala, position 140 and mutated amino acid Thr is converted to A140T. DNA mutations identified by any tool are normalized to the format “c.[position][wild type nucleotide]>[mutated nucleotide]”. In the case of insertion and deletions, given position ranges, hyphens are replaced by the underscore character (e.g. c.597-598delGA to c.597_598delGA).\n\nWe ignored EMU’s Genome category since genome variants do not appear in COSMIC or InSiGHT. We did not filter out mutation mentions based on sequence validation, so we consider all the extracted mutations from EMU. When EMU identifies a dbSNP identifier, the dbSNP API is queried to obtain further details about the mutations, identifying all available candidates for DNA and protein mutations associated with each ID. There were some mentions in which the position of the DNA or protein substitution mutation was provided as exon/intron number or a codon position. The codon positions were converted to the three candidate nucleotide positions. Exon and intron mentions were removed since no precise position could be derived.\n\nIn the curated databases, the mutations are linked to the genes or proteins where they happen. In addition to the extraction of mutations, we have annotated and normalized the genes in the documents based on a dictionary developed from the NCBI Gene database, using only the human genes. We followed the procedure in Jimeno Yepes (2013)34 and removed duplicates and filtered out certain misleading or ambiguous gene names, such as those ending with disease, syndrome, or susceptibility, and removed terms from a standard stopword list. Based on observations from previous work12, we have added the following variations to the genes related to the InSiGHT database. These terms are variations of the original gene term but prefixing the letter h to indicate that it is a human gene. Thus we have added hMSH2 for MSH2, hMSH6 for MSH6 and hPMS2 for PMS2.\n\nWe used our own dictionary because this has shown to be effective34,35 and human genes are not as ambiguous compared to other species. Since our objective is to investigate the coverage of current approaches when dealing with the curation of existing databases, we can have more control on the false positives and false negatives. In addition, we are considering resources in addition to MEDLINE citations and full text documents, for which current methods based on machine learning approaches do not need to perform as expected. We have used ConceptMapper36 (http://uima.apache.org/d/uima-addons-current/ConceptMapper/ConceptMapperAnnotatorUserGuide.html) as the dictionary tagger tool reusing the configuration prepared for the BioCreative 2013 CTD track35, which does not make case distinction, tokens have to be matched in the same order and only the longest match is considered and tokens must be adjacent to each other. The identified genes are related to the mutations based on document co-occurrences.\n\n\nResults\n\nWe have performed two types of evaluation. An intrinsic evaluation of the variant extraction methods on an annotated corpus developed for the purpose of variant curation and an extrinsic evaluation based on the ability of the methods to recover the mutations from the articles.\n\nIntrinsic evaluation of the mutation extraction tools is shown in Table 3 in terms of precision, recall and F1 score. The results are estimated based on two matching schemas: exact matching, in which the annotated entities must match exactly the span of the entities in the reference set, and partial matching, in which the annotated entities may have any overlap with the entities in the reference set. The partial matching relaxes annotation boundaries, so entities with differences like DNA or protein variant prefixes c. and p. respectively are not considered as errors.\n\nResults for exact and partial matching are present. Each row shows the performance of each method in terms of true positives (TP), false negatives (FN), false positives (FP), Precision, Recall and F1 measure (F1). The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH) and their combination (Combined).\n\nConsidering the partial matching, the precision of each tool is over 90%, which is in agreement with previously reported work on different corpora. On the other hand, the tools show different recall values. EMU and SETH show the best performance, with SETH showing better results than EMU in exact matching and in partial matching. MutationFinder has a significantly lower coverage due to its focus on point mutations. tmVar and OMM show similar partial matching performance, even though OMM extracts only protein variants. tmVar has lower performance than expected based on its previously reported performance. This might be due to the fact that tmVar was trained on abstracts, while the Variome corpus consists of full text articles. If we combine the annotations of the tools by simple merge, when evaluating the results based on partial matching, the performance is higher than any single system (precision = 0.9725, recall = 0.9060 and F1 = 0.9381), in particular due to an increase in recall.\n\nThe Variome corpus contains not only abstracts but also full text. These results show that mutation extraction tools that were developed based on MEDLINE abstracts, once applied to narrative literature still have high precision, and that their combination provides a high precision and recall solution.\n\nSince Open Mutation Miner and MutationFinder explicitly only deal with protein mutations, we have divided the results into DNA and protein mutation subsets and estimated recall for each subset, shown in Table 4 and Table 5 respectively. Unsurprisingly, OMM and MutationFinder did not recover any DNA mutation. The result on protein mutations show that MutationFinder has a very low performance, due to its coverage of only point mutations. Open Mutation Miner has a high recall, over 96%, as well as high precision as previously reported. tmVar has quite a high recall in the protein mutation set compared to the DNA mutation set. SETH has the overall highest performance, but its recall is below EMU for protein mutations. This is because many mutations in text do not exactly follow the HGVS nomenclature. Considering DNA mutations, we find that except for SETH, the performance of the other tools is lower. This is explained because there are specific types of DNA variants, e.g. deletions, that are not as well covered by the other tools as by SETH. tmVar performance is low for DNA mutations compared EMU and SETH, with a recall of 9.62%. As can be seen in Table 5, tmVar has much stronger coverage of protein mutations as compared to DNA mutations.\n\nResults for exact and partial matching are present. Each row shows the performance of each method in terms of true positives (TP), false negatives (FN) and Recall. The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nResults for exact and partial matching are present. Each row shows the performance of each method in terms of true positives (TP), false negatives (FN) and Recall. The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nTable 6 and Table 7 contain the frequency of false positives and false negatives made by each tool, grouped by type of genetic variation. Types of genetic variation were manually annotated. There are 8 DNA deletions, 2 DNA insertions/deletions, 42 DNA substitutions and 66 protein substitutions. In addition to a substitution that EMU identified incorrectly, all two other false positives are annotations that should be corrected in the Variome corpus. Specifically a protein substitution (V600E) and a DNA deletion (c.2700_2701delTC). EMU false negatives include deletions c.3927_3931del AAAGA, substitutions such as c.1852_1853AA>GC and mentions surrounded by parentheses. Considering the false negatives, MutationFinder failed at extracting mutations with three-letter amino acids (e.g., p.Pro622Thr) or single-letter amino acids without the p. prefix as M23A. OMM has only two false negatives that are protein mutations that appear within parentheses in text. tmVar has many DNA false negatives, which as indicated before, shows the low coverage of this variant type. SETH fails with noncompliant HGVS mutations, e.g. C1668 C > T that should be c.1668C > T. All DNA mutation extraction tools fail with some expressions like codon 41: A→G, which might require additional regular expressions.\n\nOnly the tools that perform DNA variant extraction are considered. The tools are Extractor of Mutations (EMU), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nThe tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nResults on the mutation extraction are available in Table 8 for the COSMIC database and in Table 9 for the InSiGHT database. The tables show results for different representations of the articles and different mutation extraction tools used. For each representation group, there is an additional row showing the result of combining the mutations extracted by each method. Each row shows the total number of mutations in the reference set, the number of mutations matched and the recall, which is just the proportion of the matched mutations with respect to the mutations in the reference set. Matching requires matching the complete triple {PMID, gene, mutation}.\n\nThe table shows the number of variants in the reference set (Total), the number of matched variants by the mutation extraction tool (Matched), the proportion of matched variants (Recall), the number of variants matched when the gene is not considered (M NG) and the proportion of matched variants when the gene is not considered (Rec NG). The data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), PDF representation for all the articles (pdf.all), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all). The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH). The row with tool value as All indicates the result when the variants extracted by all the tools are merged.\n\nThe table shows the number of variants in the reference set (Total), the number of matched variants by the mutation extraction tool (Matched) and the proportion of matched variants (Recall). When relaxing the gene matching (M NG), the results do not change, thus this data is not shown. The data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), PDF representation for all the articles (pdf.all), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all). The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH). The row with tool value as All indicates the result when the variants extracted by all the tools are merged.\n\nFor the COSMIC database, most of the mutations are found in the supplementary material, while for InSiGHT we find that most of the mutations are spread between tables and supplementary material. Low recall is obtained from the mutations extracted from the articles’ abstracts and full text representations. This is in accordance with previously published work11,12, in which a similar effect is observed.\n\nWe explored the effect of combining the tested tools together, since it is apparent several have complementary scope. The combination was implemented simply by merging the results of the systems. The combination of all systems improves the previously published text mining coverage. Coverage increases from 45.63% recall to 62.30% in the case of the InSiGHT database and from 62.30% recall to 70.56% in the case of the COSMIC database.\n\nThe coverage of the COSMIC database is larger than the InSiGHT database. In InSiGHT, a large proportion of the information is found in tables in expressions that involve an intron position, i.e. IVS17(+5)G>C, which are not properly identified by the mutation extraction systems.\n\nIn addition to these results, we have relaxed the gene matching requirements, so only the PMID and mutation are required to match. In the result tables, these results are shown in the fields “M NG” (NG=No Gene) for the number of matched mutations and the “Rec NG” for the proportion of mutations covered from the reference set. We find that there is just a small increase in the case of the COSMIC database. There is no difference for the InSiGHT database and hence this additional data is not shown for the InSiGHT database. The InSiGHT database focuses on only four genes and the dictionary seems to cover all their possible gene name variations.\n\nConsidering the tools individually, MutationFinder has low coverage of the curated mutations. This result follows the observations from the intrinsic evaluation performed on the Variome corpus. OMM is focused on protein mutations, thus it also suffers from lower recall. Given its excellent performance in intrinsic protein variant extraction, this suggests that when its performance is low compared to other methods it means that DNA variants are more common, for instance in the supplementary material in the InSiGHT database. The coverage of SETH depends largely on the compliance with HGVS nomenclature and explains why in some cases its recall is low compared to other methods. EMU provides a more robust coverage of mutations overall. However, the combination of different methods shows an increase compared to previous published work based only on this tool. This is partially explained by the coverage provided by OMM of protein mutations but also by the performance of SETH in the extraction of more complex deletions from the InSiGHT database.\n\nDuring the analysis of the results, we realized that in a small number of cases PMC makes reference to the tables of the article without including their content. In the InSiGHT database this happens only with the PMID:12373605. To mitigate this problem we downloaded all the PDF files for all the PMC documents and converted them to plain text. This is given as pdf.all in the result tables. We find that this set contains more mutations than the full text or the table sets, because full text and tables are contained in the PDF of the articles.\n\nThere are articles for which no mutation extraction tool could recover any mutation. We have performed an analysis of the coverage of the mutation extraction tools for the articles in which at least one mutation can be identified by any mutation extraction tool and at least one mutation is in the reference set, referred to as the common set. The results on the common set are available from Table 10 and Table 11. Generally the coverage of the common set is higher. This difference is most dramatic when considering the citations for which mutations can be identified in the abstract, with 23% and 29% recall in COSMIC and InSiGHT respectively.\n\nThe table shows the number of common articles (PMIDs), the number of variants in the reference set (Total), the number of matched variants by the mutation extraction tool (Matched), the proportion of matched variants (Recall), the number of variants matched when the gene is not considered (M NG) and the proportion of matched variants when the gene is not considered (Rec NG). The data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), PDF representation for all the articles (pdf.all), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all). The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH). The row with tool value as All indicates the result when the variants extracted by all the tools are merged.\n\nThe table shows the number of common articles (PMIDs), the number of variants in the reference set (Total), the number of matched variants by the mutation extraction tool (Matched) and the proportion of matched variants (Recall). When relaxing the gene matching (M NG), the results do not change, thus this data is not shown. The data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), PDF representation for all the articles (pdf.all), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all). The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH). The row with tool value as All indicates the result when the variants extracted by all the tools are merged.\n\nWhen considering tables, the coverage of COSMIC seems to increase only slightly compared to the InSiGHT database. This might mean that many of the mutations are in tables in the InSiGHT database, while this is not the case for the COSMIC database.\n\nTable 12 and Table 13 show the overlap of the mutations extracted by each tool using all the data sources from the articles. The results in the tables show the complementarity of the mutation extraction tools. MutationFinder has the lowest overlap with the mutations extracted by other systems, while EMU has the best coverage. The overlap of MutationFinder and OpenMutationMiner with other tools is lower in the InSiGHT database, which might indicate that there are proportionately more DNA mutations in this set compared to COSMIC.\n\nThe tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nThe tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nIntrinsic results show that Open Mutation Miner and SETH have the best performance for protein and DNA mutations respectively. Table 11 shows that the combination recovers a large proportion of the mutations for the COSMIC database. On the other hand, the tools still fail to identify some genetic variants, mainly when they do not follow the HGVS format, as is the case in the supplementary material in InSiGHT. When we add EMU, as shown in Table 15, the coverage increases for InSiGHT, being close to the coverage obtained by combining the different tools.\n\nThe data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all).\n\nThe data sets considered are MEDLINE abstracts (medline), Open Access PMC articles (pmc.ft), PDF articles when no Open Access PMC articles are available (pdf), tables available from the Open Access PMC Articles’ XML (table), supplementary material (sup) and the combination from all the sources (all).\n\n\nDiscussion\n\nThe Results section presented results for two types of experiments. The first one compares the coverage of current mutation extraction tools on a data set intended to support the curation of the InSiGHT database. The second type of experiment looks at the performance of the tools in the context of mutation database curation.\n\nThe result on mutation extraction shows that performance is quite high using partial matching, with a result over 85 in F1 measure by SETH. We see a big change from the performance of the mutation extraction tools between exact and partial matching regimes, mainly due to the differences in boundaries due to annotation of the prefixes “c.” and “p.” in the gold standard, while not included by the extraction tools.\n\nThe split between DNA and protein variants shows the varying scope of the tools. OMM shows a high recall of protein mutations, only missing the mentions M23A and V600E, which were sourrounded by parentheses. MutationFinder missed, in addition to the OMM examples, protein variants using three letter amino acid abbreviations, e.g. p.Pro622Thr. EMU missed protein variants surrounded by parenthesis too, and mutations for which the amino acid substitution was specified by an X, e.g. p.Arg226X. tmVar only missed protein variants surrounded by parenthesis. SETH missed any mention not following the HGVS nomenclature strictly, including variants without the p. protein variant prefix.\n\nConsidering the DNA variant annotation performed by EMU, tmVar and SETH, we find that all the tools missed expressions such as codon 33: C>A, expression that resembles the protein variants A1796 and very complex expressions. For example, c. 423 -6delAAATAGGTinsGAAGCAAGATCAG in PMID:18433509. EMU missed DNA variants that seem more complex, with ranges in the substitution c.1852_1853AA>GC or other expressions c.4236del8ins13. tmVar missed a large number of the DNA mutations. Many were substitutions with spaces within the text, e.g. c.5465A > T, expressions in parenthesis and verbose expressions, e.g. 1799T to A. SETH, in addition to the examples mentioned above, missed expressions that do not exactly follow the HGVS nomenclature, e.g. C1668C > T, even though it recovered the larger set of DNA variants.\n\nFurthermore, the remaining mutation entities without change and location information include terms that are related to mutation. In some cases, there are terms that denote the location of the mutation but not the specific change, e.g. exon 10, mention the change without specifying the position, e.g. G:C to A:T transition, name a mutated gene APC+, and terms that describe the type of mutation somatic mutation. The variety of information covered by these terms might require the use of different techniques for each type. The first three types could be annotated using more general regular expressions, while a dictionary approach might be suitable for the terms describing the mutation types.\n\nWe processed the mutation types with the NCBO annotator37 (http://bioportal.bioontology.org/annotator), which uses a large set of terminologies and ontologies including the NCI thesaurus38 and the Sequence Ontology39. Some of the terms are properly annotated with concepts from these resources, e.g. somatic mutation or germline mutation while others are not covered by any of these resources, e.g. truncating mutation, including resources such as the Unified Medical Language System (UMLS)40. Some terms are partially annotated but some simple rules could be considered to fully annotate them, for example, pathogenic mutation where pathogenic and mutation are annotated with different concepts or large mutation where only mutation is annotated.\n\nResults on the recovery of curated mutations s how two things. First, recall of curated mutations from the narrative part of the documents is very low. Second, a large number of them can be recovered from tables and supplementary material. These results echo our previous results obtained using the EMU system alone11,12, and demonstrate that the complete set of material associated with a publication is commonly considered in curation of mutation information.\n\nThere are mutations not extracted by EMU that have been found by other mutation extraction methods. Most of the annotations missed by EMU are deletions and duplicates that seem to be partly covered by EMU but are better covered by tools like SETH. There are many protein mutations missed by SETH, because of the lack of an explicit p. prefix, that are covered by Open Mutation Miner.\n\nThe most significant increase in recall of the COSMIC database happens in the supplementary material. The recall of the combination increases from 0.5671 to 0.6730. The overall recall, which was around 0.52 for EMU alone, increases to 0.7053 for the combined outputs.\n\nIn the InSiGHT database, the most significant increase in recall takes place when adding the mutations extracted from the tables. The main reason for this is the SETH tool’s identification of deletions in the tables. The overall recall, which was previously around 0.45 using EMU, increases to 0.6230 when combining the output of all the mutation tools.\n\nWe have annotated the variants available in each of the databases with a mutation type, using SETH. SETH, as mentioned in the methods section, annotates mutations based on grammar defined for HGVS22 and produces a mutation type for each string it recognizes. The list of mutation types is: SUBSTITUTION, DELETION, DELETION_INSERTION, DUPLICATION, INSERTION, FRAMESHIFT. To this set, we have added the type UNKNOWN in the case SETH does not identify a specific mutation type, typically caused by underspecified phrases such as c.? and p.?. For a few cases like the substitution p.H776_C777>QS SETH also does not return any mutation type. There are many cases in which a DNA mutation cannot be mapped to a protein mutation and frameshift mutations like L280FfsX4 are not covered by SETH.\n\nThe mutations in the InSiGHT and COSMIC databases are already in HGVS format, and so SETH can be used directly to classify each mutation in the databases by variant type. The analysis by type is available in Table 16 and Table 17. We can see that a large proportion of the variants are substitutions. The analysis of missed variants by type is available in Table 18 and Table 19.\n\nVariants have been annotated using SETH, thus the variant types delivered by SETH are considered.\n\nVariants have been annotated using SETH, thus the variant types delivered by SETH are considered.\n\nMissing variants have been annotated using SETH, thus the variant types delivered by SETH are considered.\n\nMissing variants have been annotated using SETH, thus the variant types delivered by SETH are considered.\n\nIn COSMIC, the most common type of missed mutations are DNA substitutions, mostly from two papers. These articles are PMID:21720365 with 5,589 variants and PMID:18772890 with 1,112 variants. The variants appear spread within supplementary material and in tables. We had already observed this previously11, although here we perform more detailed annotation of the entities. Most of the DNA substitutions result in a known protein mutation, although there are around 805 mutations for which the effect on the protein is unknown.\n\nIn contrast to the large number of substitutions missing from the COSMIC database, DNA deletions are the most common variant type missing in the InSiGHT database. In some cases this is due to the failure of the text mining approaches. For instance, in PMID:15655560, only the substring as 1705delAG of the deletion c.1704_1705delAG is identified by the combination of text mining tools. In addition, many of the deletions are not in a form usually expected by the mutation tools. This is the case of deletions expressed in non-standard nomenclature, e.g. del exon 3, as well as substitutions or deletions that require transformations, such as IVS17(+5)G>C, found in a table in PMID:14970868. The mutation extraction tools would need to take a closer look at these examples and incorporate the appropriate patterns.\n\nThere are mutations that are extracted by the mutation tools that do not appear in the curated databases, as found by Schench et al.29. This is because these mutations are not of the interest to the databases. This also explains the low number of matches between InSiGHT and COSMIC within the 40 overlapping articles. COSMIC focuses on somatic mutations while InSiGHT collects germline mutations related to Lynch Syndrome for just four genes. In addition, some of the extracted mentions are not functional or significant for the disease, as previously described11. For instance in PMID:10469011, the mutation Ala140Thr is extracted but the article states this mutation … is known to be functionally silent, and hence was excluded from the database.\n\nWe have looked at the types of mutations annotated by each tool, using SETH to identify the mutation type. These are available from Table 20 and Table 21. Similarly to existing results, most of the mutations are substitutions and in lower number, deletions and duplications. The tools do not reliably identify insertions, although a large number of variants that could not be annotated by SETH, and were labelled as UNKNOWN, are actually insertions. As expected, OMM and MF do not annotate DNA variants and annotate protein substitutions.\n\nThe type of each variant has been annotated using SETH, thus the variant types delivered by SETH are considered. The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH). A large number of DNA mutations are annotated by EMU. 1887448 of the COSMIC variants are substitutions derived from the supplementary material of a single article, PMID:22622578. Most of these substitutions are provided in terms of chromosome location rather than relative to the gene.\n\nThe type of each variant has been annotated using SETH, thus the variant types delivered by SETH are considered. The tools are Extractor of Mutations (EMU), OpenMutationMiner (OMM), MutationFinder (MF), tmVar and SNP Extraction Tool for Human Variations (SETH).\n\nSupplementary material for the COSMIC database and tables for the InSiGHT database contribute a large number of mutations. On the other hand, processing these type of resources poses new challenges. Examination of the supplementary files show that many of the mutations listed in MS Excel files and MS Word files are in a tabular format, thus the problem could be reduced to mutation extraction from tables. Table processing for mutation extraction has been already explored by Wong et al.41. The scope of that work was to classify the type of information in each field in a table, based on a machine-learned model. This work could be extended to properly identify the gene associated to a mutation, even if it appears in the caption of the table, in a column header, and/or as a field in the same row as the mutation. This requires additional mechanisms to extract additional fields and post-process the extracted information in order to normalize the mutation mentions.\n\n\nConclusions\n\nWe have performed a broad assessment of the state-of-the-art performance of existing tools for extraction of genetic variants from the published biomedical literature, considering and comparing five publicly available extraction tools on two complementary evaluation tasks. We have proposed combining multiple mutation extraction tools together by merging their results, and have shown that their combination results in substantially improved recall of mutations with minimal impact on precision, providing evidence of the complementary nature of the tools.\n\nOur results show that current tools have a very good performance on the narrative parts of published articles, and demonstrate that earlier performance claims for MEDLINE abstracts extend to full text. On the other hand, the excellent extraction performance on this narrative content contrasts with substantially lower recall in the context of database curation, even when all results of all considered tools are merged together. Our results demonstrate that only a small fraction of the curated mutations are available from the narrative part of the articles, and that most of the information is available only in tables and supplementary files associated to the articles. When the tools are deployed against this additional material, we are able to substantially increase recall. This increase is particularly evident when the tools are used together.\n\nWe have further examined in detail the performance of these tools on different types of genetic variants, considering not only the distinction between protein variation and DNA variation, but also contrasting performance on different types of variation, e.g. substitutions, deletions, and insertions. We demonstrate that the coverage of different tools is quite complementary with respect to these distinctions, providing an explanation for the performance benefit obtained by merging their results.\n\nFuture work involves integrating our results into a genetic variant database curation tool. Before achieving this goal, there are several improvements to perform. As we have seen, the combination of mutation extraction tools recover a large part of the mutations curated in existing databases, and therefore combining several tools together is a viable strategy for genetic variant extraction, but there are still variants that are not covered. Our error analysis shows that a particular gap is mutations that do not include an explicit location or that imply a variation in a specific region in a gene (e.g. Del exon 3). Coverage of DNA insertions is low, and could be a particular target for improvement.\n\nFurthermore, special processing is required to recover information from tables. To address this, we plan to extend work previously done by Wong et al.41. Finally, a curation tool must consider the differing scopes of different genetic variant databases, i.e. COSMIC is interested in somatic mutations while InSiGHT is interested in germline mutations. Further extension to this work would therefore include classifying the variants into these two categories.", "appendix": "Author contributions\n\n\n\nAJ designed and carried out the experiments, participated in the development of the methods and drafted the manuscript. KV designed the experiments, participated in analysis of the results, and drafted the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nNational ICT Australia (NICTA) is funded by the Australian Government as represented by the Department of Broadband, Communications and the Digital Economy and the Australian Research Council through the ICT Centre of Excellence program.\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 the InSiGHT database curator, John-Paul Plazzer of the Royal Melbourne Hospital, for sharing the InSiGHT data and helping us to interpret the database fields. We also thank the COSMIC team for helpful details about their database. We would like to thank the developers of SETH and tmVar for making their tools available. Finally, we would like to thank Andrew MacKinlay for his support in preparing the gene normalization procedure.\n\n\nReferences\n\nHamosh A, Scott AF, Amberger JS, et al.: Online Mendelian Inheritance in Man (OMIM), a knowledgebase of human genes and genetic disorders. Nucleic Acids Res. 2005; 33(Database issue): D514–D517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClaustres M, Horaitis O, Vanevski M, et al.: Time for a unified system of mutation description and reporting: A review of locus-specific mutation databases. Genome Res. 2002; 12(5): 680–688. PubMed Abstract \| Publisher Full Text\n\nBaker CJO, René W: Mutation Mining: A Prospector’s Tale. Journal of Information Systems Frontiers. 2006; 8(1): 47–57. Publisher Full Text\n\nCaporaso JG, Baumgartner WA, Randolph DA, et al.: MutationFinder: A high-performance system for extracting point mutation mentions from text. Bioinformatics. 2007; 23(14): 1862–1865. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrallinger M, Izarzugaza JM, Rodriguez-Penagos C, et al.: Extraction of human kinase mutations from literature, databases and genotyping studies. BMC Bioinformatics. 2009; 10(Suppl 8): S1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoughty E, Kertesz-Farkas A, Bodenreider O, et al.: Toward an automatic method for extracting cancer- and other disease-related point mutations from the biomedical literature. Bioinformatics. 2011; 27(3): 408–415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNaderi N, Witte R: Automated extraction and semantic analysis of mutation impacts from the biomedical literature. BMC Genomics. 2012; 13(Suppl 4): S10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHakenberg J, Voronov D, Nguyên VH, et al.: A SNPshot of PubMed to associate genetic variants with drugs, diseases, and adverse reactions. J Biomed Inform. 2012; 45(5): 842–50. PubMed Abstract \| Publisher Full Text\n\nWei CH, Harris BR, Kao HY, et al.: tmVar: a text mining approach for extracting sequence variants in biomedical literature. Bioinformatics. 2013; 29(11): 1433–1439. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nden Dunnen JT, Antonarakis SE: Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion. Hum Mutat. 2000; 15(1): 7–12. PubMed Abstract \| Publisher Full Text\n\nJimeno-Yepes A, Verspoor K: Literature mining of genetic variants for curation: Quantifying the importance of supplementary material. Database: The Journal of Biological Databases and Curation (to appear). 2014. Reference Source\n\nJimeno-Yepes A, Verspoor K: Towards automatic large-scale curation of genomic variation: improving coverage based on supplementary material. In Proceedings of BioLINK SIG 2013. 2013; 39–43. Reference Source\n\nBamford S, Dawson E, Forbes S, et al.: The COSMIC (Catalogue of Somatic Mutations in Cancer) database and website. Br J Cancer. 2004; 91(2): 355–358. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPlazzer JP, Sijmons RH, Woods MO, et al.: The InSiGHT database: Utilizing 100 years of insights into Lynch Syndrome. Familial Cancer. 2013; 12(2): 175–180. PubMed Abstract \| Publisher Full Text\n\nVerspoor K, Jimeno Yepes A, Cavedon L, et al.: Annotating the biomedical literature for the human variome. Database (Oxford) 2013; 2013: bat019. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXuan W, Wang P, Watson SJ, et al.: Medline search engine for finding genetic markers with biological significance. Bioinformatics. 2007; 23(18): 2477–2484. PubMed Abstract \| Publisher Full Text\n\nSETH. SNP extraction tool for human variations. 2013. Reference Source\n\nNagel K, Jimeno-Yepes A, Rebholz-Schuhmann D: Annotation of protein residues based on a literature analysis: Cross-validation against UniProtKb. BMC Bioinformatics. 2009; 10(Suppl 8): S4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNagel K: Automatic functional annotation of predicted active sites: Combining PDB and literature mining. PhD thesis, University of Cambridge. 2009. Reference Source\n\nRavikumar K, Liu H, Cohn JD, et al.: Literature mining of protein-residue associations with graph rules learned through distant supervision. J Biomed Semantics. 2012; 3(Suppl 3): S2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSherry ST, Ward MH, Kholodov M, et al.: dbSNP: the NCBI database of genetic variation. Nucleic Acids Res. 2001; 29(1): 308–311. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJeroen JF, Blavier A, den Dunnen JT, et al.: A formalized description of the standard human variant nomenclature in Extended BackusNaur Form. BMC Bioinformatics. 2011; 12(Suppl 4): S5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThomas PE, Klinger R, Furlong L, et al.: Challenges in the association of human single nucleotide polymorphism mentions with unique database identifiers. BMC Bioinformatics. 2011; 12(Suppl 4): S4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFurlong LI, Dach H, Hofmann-Apitius M, et al.: OSIRISv1. 2: a named entity recognition system for sequence variants of genes in biomedical literature. BMC Bioinformatics. 2008; 9(1): 84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrtutay C, Väliaho J, Stenberg K, et al.: KinMutBase: a registry of disease-causing mutations in protein kinase domains. Hum Mutat. 2005; 25(5): 435–442. PubMed Abstract \| Publisher Full Text\n\nYip YL, Scheib H, Diemand AV, et al.: The Swiss-Prot variant page and the ModSNP database: A resource for sequence and structure information on human protein variants. Hum Mutat. 2004; 23(5): 464–470. PubMed Abstract \| Publisher Full Text\n\nHurst JM, McMillan LE, Porter CT, et al.: The SAAPdb web resource: A large-scale structural analysis of mutant proteins. Hum Mutat. 2009; 30(4): 616–624. PubMed Abstract \| Publisher Full Text\n\nJia M, Forbes SA, Beare D, et al.: Mining cancer genomes in COSMIC. In BMC Proceedings. 2012; 6(Suppl 6): 17. Publisher Full Text\n\nSchenck M, Politz O, Groth P: Extraction of genetic mutations associated with cancer from public literature. J Health Med Informat. 2012; (S2). Publisher Full Text\n\nCaporaso JG, Deshpande N, Fink JL, et al.: Intrinsic evaluation of text mining tools may not predict performance on realistic tasks. Pac Symp Biocomput. 2008; 640–651. PubMed Abstract \| Free Full Text\n\nVerspoor K, MacKinlay A, Cohn JD, et al.: Detection of protein catalytic sites in the biomedical literature. Pac Symp Biocomput. 2013; 18: 433–444. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRance B, Doughty E, Demner-Fushman D, et al.: A mutation-centric approach to identifying pharmacogenomic relations in text. J Biomed Inform. 2012; 45(5): 835–841. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVerspoor K, Cohen KB, Hunter L: The textual characteristics of traditional and Open Access scientific journals are similar. BMC Bioinformatics. 2009; 10: 183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJimeno-Yepes JA, Sticco JC, Mork JG, et al.: GeneRIF indexing: sentence selection based on machine learning. BMC Bioinformatics. 2013; 14: 171. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacKinlay A, Verspoor K: A Web Service Annotation Framework for CTD Using the UIMA Concept Mapper. BioCreative Challenge Evaluation Workshop. 2013; 1. Reference Source\n\nMichael AT, Anni C, Igor LS: The ConceptMapper Approach to Named Entity Recognition. LREC. 2010. Reference Source\n\nClement J, Nigam S, Cherie Y, et al.: NCBO annotator: semantic annotation of biomedical data. In International Semantic Web Conference, Poster and Demo session. 2009. Reference Source\n\nSioutos N, de Coronado S, Haber MW, et al.: NCI Thesaurus: a semantic model integrating cancer-related clinical and molecular information. J Biomed Inform. 2007; 40(1): 30–43. PubMed Abstract \| Publisher Full Text\n\nEilbeck K, Lewis SE, Mungall CJ, et al.: The Sequence Ontology: a tool for the unification of genome annotations. Genome Biol. 2005; 6(5): R44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBodenreider O: The Unified Medical Language System (UMLS): integrating biomedical terminology. Nucleic Acids Res. 2004; 32(suppl 1): D267–D270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWong W, Martinez D, Cavedon L: Extraction of named entities from tables in gene mutation literature. BioNLP. 2009; 46–54. Reference Source" }	[ { "id": "3233", "date": "05 Feb 2014", "name": "Philippe Thomas", "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 comprehensively evaluates a series of named entity recognition tools for identifying mutation mentions. The different tools are evaluated using two kinds of scenarios: First, they are evaluated in an intrinsic setting, using the manually annotated \"Variome\" corpus. This strategy represents the typical evaluation of named entity recognition tools. Second, tools are compared in an extrinsic context, where the goal is to reconstruct mutational information contained in the COSMIC and InSiGHT databases. In the latter experiment, the authors identified supplemental material as the biggest resource for mutation mentions. The overall verdict of this publication is that several tools have complementary coverage and combination of methods leads to a performance increase in both evaluation settings. Finally, the authors provide some ideas to improve the recognition of genetic variants in text.This work is scientifically sound and provides an interesting analysis of five different mutation recognition tools. I only have a few remarks considering the presentation: The results for the intrinsic evaluation, shown in Table 5, require some more details. For all tools a considerable increase of recall can be observed when switching from the exact matching to the partial matching scenario. However, for OpenMutationMiner, recall increases by almost 86 percentage points whereas the increase for the other tools is about 10 percentage points. This is a surprising result and should be at least briefly mentioned in the text. A frequently found statement is that SETH recognizes only mutation mentions described in HGVS nomenclature. However, as a participant in the SETH project (see competing interests), I suggest that this claim is incorrect as SETH builds on MutationFinder to detect mutations not following the nomenclature. Furthermore, SETH modifies MutationFinder's original capabilities in order to match a wider scope of mutations (DNA mutations, nonsense mutations, and ambiguous mutations) not following the HGVS nomenclature. This is done by modifying the original MutationFinder implementation together with additional and modified regular expressions. Furthermore, SETH contains a separate component (OldNomenclature.java) for the recognition of insertions, deletions, and frameshift mutations written in deprecated HGVS nomenclature. For these reasons it is surprising, that SETH achieves lower recall than MutationFinder in many extrinsic evaluation settings (Table 8 and Table 9). In my opinion, SETH should at least find as many mutations as MutationFinder. Please use the citation provided on the SETH project site to cite SETH. (Thomas, P., Rocktäschel, T., Mayer, Y., and Leser, U. (2014). SETH: SNP Extraction Tool for Human Variations.http://rockt.github.io/SETH/. )I also have a list of minor remarks, which I would like to mention here. However, these are just ideas to further improve the publication.Table 3 shows that merging results of all tools increases performance in the partial evaluation setting. It would be interesting to see how this compares to exact matching. Do you see similar effects? It would be great to mention the numbers in the table. \"For a few cases like the substitution p.H776_C777>QS SETH also does not return any mutation type. There are many cases in which a DNA mutation cannot be mapped to a protein mutation and frameshift mutations like L280FfsX4 are not covered by SETH.\" I find this example a bit contrived, as neither p.H776_C777>QS nor L280FfsX4 follow the HGVS nomenclature. The sentence \"There are many cases in which a DNA mutation cannot be mapped to a protein mutation and frameshift mutations like L280FfsX4 are not covered by SETH\" on page 13 is (in my opinion) confusing and could be reformulated. If I understand correctly, you are saying that COSMIC and InSiGHT cannot provide protein mutations for all DNA mutations. Integration of different tools usually is an ungrateful task. It would be very interesting if the authors could share some of their experiences made using the five different tools. \"The tools do not reliably identify insertions, although a large number of variants that could not be annotated by SETH are actually insertions.\" It would be highly interesting to see a list of these mutations (maybe as Supplemental data), to help tool developers improve named entity recognition tools. Additionally, I think that it would be very valuable to add all mutation mentions extracted by the five different tools to the Supplement as well. This information should facilitate the improvement of existing named entity recognition tools.", "responses": [ { "c_id": "827", "date": "20 May 2014", "name": "Antonio José Jimeno Yepes", "role": "Author Response", "response": "Thank you for your comments and remarks. You can find the answers to the points you raised below:1. Indeed, the increase in performance is significant. The reason is that some tools do not annotate the prefix of the mutation. For instance, in the corpus there is the mutation p.Val600Glu but OMM just annotates Val600Glu. This means that using exact match, all these mutations will be identified as false positives instead of true positives. We have added a statement to this effect in the text.2. The claim about non HGVS compliance has been rewritten in the manuscript - The reason for this is that we have considered the version of SETH in which only its main components are used, which does not include MutationFinder. This point has been clarified in the paper.3. This has been corrected in the manuscript.The answers to your minor remarks can be found below:1. We have updated the results in Table 3. There is an increase in recall in both cases and the results, which is a bit better than previously reported in the partial matching approach. We have corrected a problem with the tmVar input files and this explains further changes in the results.2. Thank you for pointing this out. This implies as well that a small part of the mutations in the databases are not the fully compliant with the HGVS nomenclature. This point has been made in the manuscript with the following text: On the other hand, neither p.H776_C777>QS or L280FfsX4 follow the HGVS nomenclature. This implies that there is a small portion of the mutations coded in the mutation databases are not fully compliant with the HGVS nomenclature.3. Thank you for pointing out this issue. The sentence has been rewritten as: COSMIC and InSiGHT cannot provide protein mutations for all DNA mutations thus these mutations could not be annotated by SETH. Frameshift mutations like L280FfsX4 are not covered by SETH.4. Technical Notes have been added for each tool, in a brief “Technical Notes:” section for each tool.5. A file has been added as supplementary material in a bitbucket repository (https://bitbucket.org/readbiomed/mutationtoolcomparison), within the data folder, with these mutations and has been referenced in the paper." } ] }, { "id": "3961", "date": "27 Mar 2014", "name": "Max Haeussler", "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 have done the formidable job of running five programs that search for mutations in English text on 10 annotated full text articles (they call this \"intrinsic\") and also on articles that have been annotated by curators of two databases (\"extrinsic\"). The results make sense, they analyze them well, the article is very readable and the methods are suitable. The conclusions are relevant for everyone who is developing mutation mining tools, probably most users of text mining algorithms, and they are well justified. Why I have reservations: The authors were able to conduct this study thanks to a great practice in bioinformatics, namely to provide the source code of the program used in a scientific article. They were able to download the source code of at least three programs from other websites or supplemental data and were even able to benchmark a program which has not been published yet in an article (SETH), and found that it led the field. This shows how the state of the art progresses by sharing code (even without publications) - the authors did not have to write their own mutation finding tool, they could run existing programs on a new corpus. However, it is hard to understand why the authors do not do the same. Anyone who would like to inspect or extend the results of this article would have to redo the same work again: write parsers for five programs, convert both corpora, put the main corpus into a suitable format, including downloading some files from EuropePMC, others from PMC, and write a program to evaluate the differences. I fully agree with Referee 1 (Philippe Thomas) that when claims about performance of open source programs are made that have been written by various teams, then at least the corpus + annotations have to be added as a supplemental file. Otherwise it is very hard for a field to improve their algorithms and validate claims made in the benchmark. In this case, the code for running the mutation detection programs should also be provided. Philippe Thomas' (the author of SETH) comment that the results of SETH should have been better for a particular case can only be answered by looking at the code, the version (or the github commit ID in case of pre-publication code) the authors used and the converter, to find out where the differences come from. For a reader, it is currently impossible to validate the observation about SETH. This could be really easy to change, if the authors just attached their corpus+program as a supplemental file or on github, like most other authors in this field today.", "responses": [ { "c_id": "828", "date": "20 May 2014", "name": "Antonio José Jimeno Yepes", "role": "Author Response", "response": "Thank you for your suggestions. We have made available evaluation files in a bitbucket repository (https://bitbucket.org/readbiomed/mutationtoolcomparison). The brat mutation annotation files for the Variome corpus evaluation have been included as well. In addition, the tool brateval (https://bitbucket.org/nicta_biomed/brateval) can be used to reproduce the evaluation presented in the paper for the experiments with the Variome corpus. There are three sets of annotation files. One for all the mutations and then one set for the protein mutations only and another one for the DNA mutations only.The benchmark files for the database evaluation have been made available for COSMIC and for InSiGHT. We have made available the PMIDs of the articles and the genes and mutations curated in each one of them.We have not redistributed the text of the articles since due to license restrictions it is not possible to redistribute them. They can be obtained from the providers of the documents as indicated in the paper. The code to collect the data from PMC has been provided." } ] }, { "id": "3722", "date": "28 Mar 2014", "name": "Cecilia N. Arighi", "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 shows a thorough evaluation of the performance of five text mining tools for mutation extraction. The authors perform two types of evaluation: one to test the ability of the tools in identifying specific mentions of genetic variation in the manually-annotated Variome corpus and the other, in the context of database curation, to compare results against the data in COSMIC and InSiGHT databases. The authors analyzed the tools both individually and combined. The main findings include:The tools have complementary coverage of genetic variants mentioned in the literature (e.g. some tools covered only single point mutations whereas others covered deletions and insertions as well), therefore the combination of tools proves to be more effective.The external evaluation on the database corpora is important to assess the utility of the tools for biocuration.The main results derive from the location of the data in the article.The majority is located in the Supplementary materials (for COSMIC) and also in Tables (for InSiGHT).The authors also demonstrated that the tools, even when they were developed for medline abstracts, can be applied to other text. The article is well written, provides sufficient background information, is scientifically sound, and both the title and abstract reflect the main research results and conclusions of this work. However, there is still room for improvement.Here are some suggestions:The article talks about mutations referring more narrowly to natural gene variants (at least that would be the case for the databases mentioned above). However, at least one tool (MutationFinder) seems to consider mutation in a broader sense. The fact that the corpus was based on PMIDs in PDB suggests that these are most likely to be mutations introduced by mutagenesis (e.g. I checked a couple of examples of the corpus used by MF and, in fact, some are mutagenesis experiments) rather than genetic variants. This may also affect the performance of the tool if most of the examples that it was exposed to were of these kind. I would suggest including some comment on this in the article. The description of TmVar in the background section should include the type of variant it is able to extract. This description is present in all other systems and should be described in the paragraph where they are referred to in the paper. Both exact and partial matches are used for the calculation of performance. Partial match is simply defined as “annotated entities may have any overlap with the entities in the reference set”. Have you analyzed the range of percent overlap? Tables 8, 9, 14 and 15 need some clarification. I understand that the point the authors want to convey is about the coverage not only of the tools, but of the data in the various sections of the article (main text, tables, supplementary material, abstract vs full length).The value therefore, that is presented in the “Recall” column is like a fractional recall, as it is compared to the total number of variants in the reference set. The last row is probably the only line presenting the total Recall measure. The table is missing a row before the last one with Dataset:ALL, Tool: each tool. This row and the previous one should be highlighted as these are the ones showing the total recall. I would also add a column with the number of articles and annotations in each of the data sets (medline, pmc, pdf, etc), and the performance at each of these levels, i.e. calculate the recall on a section (Medline) based on the total annotation present in such section.\n\nI would change the title of Recall for the one showing fractional recall (calculated against Total). Maybe use a term such as coverage fraction or coverage recall? (Except for the Data set all).", "responses": [ { "c_id": "829", "date": "20 May 2014", "name": "Antonio José Jimeno Yepes", "role": "Author Response", "response": "Thank you for your suggestions. You can find our answers to the points you raised below:1. This is an important detail that allows for a better understanding of the behaviour by MutationFinder. We have added the following sentence in the discussion section: In addition, MutationFinder was developed on citations available from PDB. It is quite likely that the mutations from these citations have been introduced by mutagenesis, instead of the natural gene variants covered in COSMIC and InSiGHT, and thus its performance might be influenced by this.2. Thank you for pointing this out. The following sentence has been added to the manuscript: tmVar aims to cover a wide range of sequence variants in both protein and gene levels in HGVS format.3. This is a good idea since it provides a better overview of the matched mutations. We have added the information about the character overlap in tables 3, 4 and 5. Discussion about the results has been added to the manuscript.4. A set of rows for Dataset:All have been added in tables 8 to 10. 5. This would be an interesting result. Unfortunately, the curated databases used for this extrinsic evaluation do not offer this level of detail, i.e. it is not possible to tell from which source the information was curated from (i.e. whether the curator used an abstract, full text, etc.).6. Since the recall per data set section cannot be obtained, we have decided to leave the current recall label." } ] } ]	1	https://f1000research.com/articles/3-18
https://f1000research.com/articles/3-124/v1	10 Jun 14	{ "type": "Opinion Article", "title": "CETP inhibitors and cardiovascular disease: Time to think again", "authors": [ "Norman E Miller" ], "abstract": "Inhibition of cholesteryl ester transfer protein (CETP) lowers plasma low-density lipoprotein cholesterol concentration and raises high-density lipoprotein (HDL) cholesterol, suggesting it might prevent cardiovascular disease (CVD). From the outset, however, the concept has been controversial owing to uncertainty about its effects on HDL function and reverse cholesterol transport (RCT). Although there has long been good evidence that CETP inhibition reduces atherosclerosis in rabbits, the first information on CETP as a CVD risk factor in a prospectively followed cohort was not published until after the first Phase 3 trial of a CETP inhibitor had begun. The worrying finding that CVD incidence was related inversely to plasma CETP has since been reproduced in each of five further prospective cohort studies. Similar results were obtained in subjects on or off statin therapy, for first and second CVD events, and for mortality as well as CVD morbidity. Additionally, two recent studies have found alleles of the CETP gene that lower hepatic CETP secretion to be associated with an increased risk of myocardial infarction. Meanwhile, CETP gene transfer in mice was found to increase RCT from peripheral macrophages in vivo, and human plasma with high CETP activity was shown to have a greater capacity to remove cholesterol from cultured cells than plasma with low activity. This mounting evidence for a protective function of CETP has been given remarkably little attention, and indeed was not mentioned in several recent reviews. It appears to show that CETP inhibition does not test the HDL hypothesis as originally hoped, and raises a pressing ethical issue regarding two Phase 3 trials of inhibitors, involving more than forty thousand subjects, which are currently in progress. As the weight of evidence now clearly supports an adverse effect of CETP inhibition on CVD, an urgent review is needed to determine if these trials should be discontinued.", "keywords": [ "Cholesteryl ester transfer protein (CETP) is a hydrophobic glycoprotein in plasma that catalyzes the transfer of neutral lipids between plasma lipoproteins1. The notion that inhibition of CETP activity might prevent coronary heart disease (CHD) was based on the knowledge that it both reduces plasma low-density lipoprotein (LDL) cholesterol concentration", "and raises high-density lipoprotein (HDL) cholesterol. While there is abundant evidence that reduction of LDL cholesterol is likely to be beneficial", "the effect of CETP inhibition on the function of HDL in reverse cholesterol transport from tissues (RCT) has been uncertain." ], "content": "Background\n\nCholesteryl ester transfer protein (CETP) is a hydrophobic glycoprotein in plasma that catalyzes the transfer of neutral lipids between plasma lipoproteins1. The notion that inhibition of CETP activity might prevent coronary heart disease (CHD) was based on the knowledge that it both reduces plasma low-density lipoprotein (LDL) cholesterol concentration, and raises high-density lipoprotein (HDL) cholesterol. While there is abundant evidence that reduction of LDL cholesterol is likely to be beneficial, the effect of CETP inhibition on the function of HDL in reverse cholesterol transport from tissues (RCT) has been uncertain.\n\nIn 1996 Fielding and Havel2 argued against a hasty commitment to CETP inhibitors, drawing attention to evidence that CETP participates in the remodelling of the cholesteryl ester-rich α-HDLs that generates the small lipid-poor preβ-HDLs that are the primary acceptors of cholesterol via the ABCA1 transporters in cell membranes3. The rise in HDL cholesterol might be misleading, they argued, and reflect only retention of cholesteryl esters in the particles, while the uptake of cholesterol from arterial cells is diminished. Nevertheless, encouraged by reports that CETP gene transfer induced atherosclerosis in mice4 and that CETP inhibition prevented atherosclerosis in rabbits5–7, drug discovery programmes made rapid progress.\n\nThe case for inhibition was weakened when later studies of CETP transgenic mice contradicted the earlier findings8,9, and the incidence of CHD was not found to be significantly reduced in familial CETP deficiency10. Studies of the relation of CHD to single nucleotide polymorphisms (SNPs) of the CETP gene yielded disparate outcomes, which were not resolved by meta-analyses11–13. When Dullaart et al.13 meta-analyzed data on the Taq1B SNP (rs708272) from population-based and high-risk groups separately, the odds ratio for cardiovascular disease (CVD) in homozygotes for the B2 allele (who have low CETP activity) was 0.84 in the high-risk subjects, but 1.45 in the population-based samples. This suggested that low CETP activity actually increases CVD risk, and that its seeming protective effect in some studies may have resulted from selection towards a lower frequency of the B2 allele in high-risk groups.\n\nThe dual uncertainties over the effect of CETP inhibition on HDL function and whether it is more likely to reduce or increase CVD in humans were unresolved when the first Phase 3 study (ILLUMINATE; NCT00134264) of a member of this class of drugs (torcetrapib) was started in July 200414. In June 2007, the trial was terminated after it had become clear that the treatment had increased the incidence of the primary CVD endpoint. The authors concluded that this was probably owing to an unanticipated rise in blood pressure, but three findings challenged this interpretation. CHD mortality was related inversely to increment in blood pressure; the incidence of stroke was not raised by treatment; and non-cardiovascular mortality was increased. A Phase 3 trial of dalcetrapib (Dal-OUTCOMES; NCT00658515), a less potent inhibitor with little or no effect on blood pressure, was started in April 2008. In November 2012, this study was also terminated when it became evident that the outcome was not going to be positive15.\n\nIn the same year that ILLUMINATE enrolled its first participant, the first prospective data on plasma CETP as a risk factor for clinical CVD events were also published. Boekholdt et al.16 had found that CETP concentration, with which CETP activity is linearly related17, did not differ significantly between 1,400 controls and 735 subjects who had developed myocardial infarction (MI) during six years of follow-up, although a positive association was seen on post hoc analysis in subjects with plasma triglycerides exceeding the median of 1.7 mmol/l.\n\n\nRecent findings on plasma CETP as a CVD risk factor\n\nThe first prospective observational cohort study of plasma CETP activity or concentration as a risk factor for clinical CVD events did not appear until July 2006, two years after ILLUMINATE had started. Since then five further similar studies have been published. All six studies found CVD incidence to be related inversely to plasma CETP18–23. The designs and results of these studies are summarised in Table 1. Cohorts ranged from 1,002 to 3,256 subjects, and follow-up periods from two to 15 years (weighted average, 7.6 years). One study followed men and women separately23. Three were of first CVD events in healthy subjects18,19,23; two were of second events in subjects with an existing history of CHD20,21; and in one study primary and secondary events were pooled22. In two studies that looked at mortality in addition to CVD morbidity21,22 this was also negatively associated with CETP. Results in subjects taking pravastatin or atorvastatin18,20 mirrored those in other subjects. The suggestion in the earlier case-control study16 that subjects with raised triglycerides might differ from others was not confirmed.\n\nStudies limited to coronary angiography without clinical endpoints are not included. AMI, acute myocardial infarction; BMI, body mass index; CAD, coronary artery disease; CHD, coronary heart disease; CVD, cardiovascular disease; HR, hazard ratio; OR, odds ratio; PLTP, phospholipid transfer protein.\n\n\nRecent genome-wide analyses of CETP alleles and CVD\n\nThe CETP gene has been mapped to locus 16q21. It spans about 25 kb, and consists of 16 exons and 15 introns. In the absence of a clear picture from candidate gene studies of the association of SNPs with CVD, two genome-wide analyses have recently been published, whose results appeared to conflict with those of the observational epidemiology. In a study of more than 350,000 SNPs in 18,245 women followed for 10 years, Ridker et al.24 observed that three SNPs in or around the CETP gene (rs708272, rs4329913, rs7202364) were associated with increased HDL cholesterol and a reduced incidence of MI. A subsequent Mendelian randomization analysis found a single SNP of CETP (rs3764261) to be associated with raised HDL cholesterol and an apparent four per cent reduction in the incidence of MI25.\n\n\nReconciling the observational epidemiology and genome-wide analyses\n\nThe reliability of prospective observational epidemiology for the identification of causal effects in complex diseases has been in the spotlight of late, after some results were not confirmed in randomized clinical trials. Could the results of the recent observational studies of CETP be another instance of confounding or reverse causation? Confounding seems unlikely as multivariate analyses found the relation between CVD events and CETP concentration or activity to be independent of age, gender, hypertension, body mass index, plasma triglycerides, adiponectin, diabetes, and smoking habit18–23. Reverse causation due to a reduction of plasma CETP in response to vascular inflammation also seems improbable, as the association persisted after adjustment for plasma homocysteine, interleukin-6 and C-reactive protein concentrations22,23.\n\nAs discussed by several authors26–28, genome-wide analyses are also not without their limitations, and several aspects of the two studies warrant consideration. One is that there appears to have been no concordance between them in the alleles found to be associated with MI. Second, as data on plasma CETP were not available to either study, the relations with disease could have been owing to linkage with other genes that affect HDL and MI through independent mechanisms. The strongest association in the first study24 was with rs708272, the Taq1B SNP of CETP. This intronic polymorphism has no direct affect on CETP activity. Furthermore, the allele associated with low incidence of MI has also been found to be associated with a low prevalence of metabolic syndrome29, a potential confounder being itself a strong risk factor for CVD. The other two alleles were remote from the CETP gene, being in SLC12A3 and NUP93, respectively a solute transporter gene and the gene for a nucleoporin. The relation of MI to rs3764261 described in the more recent genome-wide study was adjusted for age and sex, but not for other potential confounders25. Furthermore, the result has since been contradicted by a meta-analysis of data from 16,570 subjects30, which found the T allele of the same SNP to be associated with reduced effectiveness of statins in preventing MI.\n\nPapp et al.31 have recently addressed the issue of genetically determined low CETP activity by using mRNA allelic expression and splice isoform assays to identify genetic variants that affect plasma CETP concentration, and then examining their relation to incident MI. In studies of 94 human livers, a common alternatively spliced isoform lacking exon 9 prevented CETP secretion in a dominant-negative manner. Increased formation of this isoform was exclusively associated with two polymorphisms in high linkage disequilibrium: one in exon 9 of CETP (rs5883-C>T), which alters an exonic splicing enhancer sequence, and another in intron 8 (rs9930761-T>C), which changes a splicing branch point nucleotide. In the INVEST-GENES prospectively followed cohort, it was found that rs5883T/rs9930761C were associated with high incidence rates of MI and stroke (P = 0.005) despite raised HDL cholesterol, strongly reinforcing the observational epidemiology.\n\n\nRecent studies of CETP, HDL function and reverse cholesterol transport\n\nWhile the epidemiologic landscape has thus evolved, laboratory research has strengthened the evidence that CETP plays an important role in RCT. Tanigawa et al.32 found that hepatic CETP gene transfer in mice stimulated the transport of cholesterol from peritoneal macrophages to the liver, followed by its elimination as bile acids. Tchoua et al.33 independently confirmed this result, and showed that the effect was blocked when the animals were given torcetrapib. There is no accepted method for quantifying reverse cholesterol transport in vivo in humans, but three groups have recently reported that human plasma with high CETP activity had a greater capacity to promote cholesterol efflux from cultured cells than plasma with low activity34–36. Villard et al.34 showed further that addition of purified CETP increased both the preβ-HDL concentration in normal human plasma and its capacity to remove cholesterol from cultured cells, reproducing an earlier result obtained with plasma from a subject with familial CETP deficiency37. Thus, the confusion over the contribution of CETP to RCT appears to have been resolved, and the concerns expressed by Fielding and Havel almost 20 years ago substantiated.\n\n\nPerspective\n\nThe history of the hypothesis that CETP inhibition will prevent atherosclerosis can be summarised thus. At the outset, our understanding of HDL biochemistry did not permit any predictions of its effect on RCT, but was sufficient to tell us that it might go either way. In the absence of information on the relation of CVD risk to CETP activity in humans, enthusiasm for the concept was fuelled by positive results in cholesterol-fed rabbits, which seemed to confirm that a rise in HDL cholesterol is a dependable biomarker of benefit. However, the first prospective cohort study of CETP as a CVD risk factor challenged this assumption. Since then, five further prospective observational studies have left no doubt that in populations CVD risk is related inversely to CETP activity. Confounding and reverse causation seem unlikely explanations. Although two genome-wide analyses appeared to have produced contrary evidence, for the reasons discussed they have not refuted the observational data. On the other hand, the latter have been reinforced by reports that subjects with functional CETP alleles that lower CETP secretion have an increased risk of MI. Thus, the weight of evidence has now shifted to the likelihood that CETP inhibition will have an adverse effect on CVD outcomes, not the beneficial effect that was hoped for. Recent laboratory studies on the impact of CETP activity on the cholesterol transport function of HDL have been consistent with this interpretation.\n\nThis interpretation does not conflict with the anti-atherogenic effect of CETP inhibition in rabbits. Apart from the obvious possibility of a species specific difference in cholesteryl ester dynamics, Shimoji et al.38 reported that dalcetrapib increases the synthesis rate of the major HDL protein (apo AI) in rabbits by 44 per cent, an effect that on its own would be expected to substantially reduce atherosclerosis39. By contrast, inhibition of CETP with torcetrapib had no effect on apo AI synthesis in humans40. It is also worth noting that probucol, which increases CETP activity41, also prevents atherosclerosis in rabbits despite lowering apo AI synthesis rate41,42.\n\nAlthough the body of disquieting data has been growing for several years, there has been surprisingly little public discussion of the issue. The paper describing the outcome of ILLUMINATE14 made no reference to the results of Marschang et al.18 published the year before. Likewise, the report on Dal-OUTCOMES15 made no mention of any one of the six observational cohort studies listed in Table 1, all of which were already in print. The same is true of an article investigating the harm caused by torcetrapib in ILLUMINATE43, and of several recent review articles44–48.\n\n\nImplications\n\nThese recent developments have significant implications. First, they are consistent with other evidence that plasma HDL cholesterol concentration is not a reliable marker of the efficiency of RCT49,50. Second, they show that clinical trials of CETP inhibitors do not test the HDL hypothesis in the manner originally envisaged. Third, they raise a pressing issue in the context of two Phase 3 studies of second generation CETP inhibitors currently in progress. ACCELERATE (NCT01687998)51, which began in 2012 and is expected to finish in 2016, has enrolled about 12,000 patients with high-risk CVD to assess the efficacy of evacetrapib52 in preventing CVD events. REVEAL (NCT01252953)53, commenced in 2011 and expected to be completed in 2017, has enrolled 30,624 patients for a similar study of anacetrapib54. In both studies, the patients in each arm are being given a statin to control LDL concentration prior to randomization.\n\nAnacetrapib is the most potent CETP inhibitor to date, and was found in DEFINE (NCT00685776)54 to lower LDL cholesterol by 50 per cent compared with the 25 per cent achieved with torcetrapib14. It is theoretically possible that this greater impact on LDL will override any adverse effect on HDL function, but it is equally possible that its greater impact on HDL cholesterol (140 per cent increase compared with 70 per cent) reflects such an extreme disturbance of HDL metabolism that its consequences will predominate. Neither the prospective epidemiology nor studies of familial CETP deficiency have provided evidence of a fall in CVD risk at extremely low activities. Although DEFINE recorded no increase in CVD in patients given anacetrapib, the authors noted that the study was too small to provide reliable information on clinical events54.\n\n\nConclusion\n\nGiven that the tide of evidence has turned so strongly against CETP inhibition in recent years, the question must be asked of whether it is now ethical to continue with the two Phase 3 trials in progress. A clinical trial is considered to be ethical only if it has a sound scientific basis and a favourable risk-benefit balance55. The two trials in question no longer satisfy either requirement, as there is clearly a strong possibility that the drugs will have exactly the opposite effect on CVD to that intended. Some might argue that there is no cause for concern, as morbidity and mortality are being regularly reviewed by data monitoring committees. However, such committees can intervene only when pre-specified statistical criteria have been met, by which time many participants may have suffered harm.\n\nCarrying on and hoping for the best is not an acceptable option. An independent review is urgently needed to determine if the trials should be discontinued.", "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\nReferences\n\nFielding CJ, Fielding PE: Molecular physiology of reverse cholesterol transport. J Lipid Res. 1995; 36(2): 211–28. PubMed Abstract\n\nFielding CJ, Havel RJ: Cholesteryl ester transfer protein: friend or foe? J Clin Invest. 1996; 97(12): 2687–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHennessy LK, Kunitake ST, Kane JP: Apolipoprotein A-I-containing lipoproteins, with or without apolipoprotein A-II, as progenitors of pre-beta high density lipoprotein particles. Biochemistry. 1993; 32(22): 5759–65. PubMed Abstract \| Publisher Full Text\n\nMarotti KR, Castle CK, Boyle TP, et al.: Severe atherosclerosis in transgenic mice expressing simian cholesteryl ester transfer protein. Nature. 1993; 364(6432): 73–5. PubMed Abstract \| Publisher Full Text\n\nOkamoto H, Yonemori F, Wakitani K, et al.: A cholesteryl ester transfer protein inhibitor attenuates atherosclerosis in rabbits. Nature. 2000; 406(6792): 203–7. PubMed Abstract \| Publisher Full Text\n\nSugano O, Makino N, Sawada S, et al.: Effect of antisense oligonucleotides against cholesteryl transfer protein on the development of atherosclerosis in cholesterol-fed rabbits. J Biol Chem. 1998; 273(9): 5033–6. PubMed Abstract \| Publisher Full Text\n\nRitterhaus CW, Miller DP, Thomas LJ, et al.: Vaccine-induced antibodies inhibit CETP activity in vivo and reduce aortic lesions in a rabbit model of atherosclerosis. Arterioscler Thromb Vasc Biol. 2000; 20(9): 2106–12. PubMed Abstract \| Publisher Full Text\n\nHayek T, Masucci-Magoulas L, Jiang X, et al.: Decreased early atherosclerotic lesions in hypertriglyceridemic mice expressing cholesteryl ester transfer protein transgene. J Clin Invest. 1995; 96(4): 2071–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFöger B, Chase M, Amar MJ, et al.: Cholesteryl ester transfer protein corrects dysfunctional high density lipoproteins and reduces aortic atherosclerosis in lecithin cholesterol acyltransferase transgenic mice. J Biol Chem. 1999; 274(52): 36912–20. PubMed Abstract \| Publisher Full Text\n\nCurb JD, Abbott RD, Rodriguez BL, et al.: A prospective study of HDL-C and cholesteryl ester transfer protein gene mutations and the risk of coronary heart disease in the elderly. J Lipid Res. 2004; 45(5): 948–53. PubMed Abstract \| Publisher Full Text\n\nThompson A, Di Angelantonio E, Sarwar N, et al.: Association of cholesteryl ester transfer protein genotypes with CETP mass and activity, lipid levels, and coronary risk. JAMA. 2008; 299(23): 2777–88. PubMed Abstract \| Publisher Full Text\n\nLi YY, Wu XY, Xu J, et al.: Apo A5 -1131T/C, FgB -455G/A, -148C/T, and CETP TaqIB gene polymorphisms and coronary artery disease in the Chinese population: a meta-analysis of 15,055 subjects. Mol Biol Rep. 2013; 40(2): 1997–2014. PubMed Abstract \| Publisher Full Text\n\nDullaart RP, Sluiter WJ: Common variation in the CETP gene and the implications for cardiovascular disease and its treatment: an updated analysis. Pharmacogenomics. 2008; 9(6): 747–63. PubMed Abstract \| Publisher Full Text\n\nBarter PJ, Caulfield M, Eriksson M, et al.: Effects of torcetrapib in patients at high risk for coronary events. N Engl J Med. 2007; 357(21): 2109–22. PubMed Abstract \| Publisher Full Text\n\nSchwartz GG, Olsson AG, Abt M, et al.: Effects of dalcetrapib in patients with a recent acute coronary syndrome. N Engl J Med. 2012; 367(22): 2089–99. PubMed Abstract \| Publisher Full Text\n\nBoekholdt SM, Kuivenhoven JA, Wareham NJ, et al.: Plasma levels of cholesteryl ester transfer protein and the risk of future coronary artery disease in apparently healthy men and women: the Prospective EPIC (European Prospective Investigation into Cancer and Nutrition)–Norfolk population study. Circulation. 2004; 110(11): 1418–23. PubMed Abstract \| Publisher Full Text\n\nRitsch A, Auer B, Foger B, et al.: Polyclonal antibody-based immunoradiometric assay for quantification of cholesteryl ester transfer protein. J Lipid Res. 1993; 34(4): 673–79. PubMed Abstract\n\nMarschang P, Sandhofer A, Ritsch A, et al.: Plasma cholesteryl ester transfer protein concentrations predict cardiovascular events in patients with coronary artery disease treated with pravastatin. J Intern Med. 2006; 260(2): 151–9. PubMed Abstract \| Publisher Full Text\n\nVasan RS, Pencina MJ, Robins SJ, et al.: Association of circulating cholesteryl ester transfer protein activity with incidence of cardiovascular disease in the community. Circulation. 2009; 120(24): 2414–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhera AV, Wolfe ML, Cannon CP, et al.: On-statin cholesteryl ester transfer protein mass and risk of recurrent coronary events (from the pravastatin or atorvastatin evaluation and infection therapy–thrombolysis in myocardial infarction 22 [PROVE IT-TIMI 22] study). Am J Cardiol. 2010; 106(4): 451–6. PubMed Abstract \| Publisher Full Text\n\nDuwensee K, Breitling LP, Tancevski I, et al.: Cholesteryl ester transfer protein in patients with coronary heart disease. Eur J Clin Invest. 2010; 40(7): 616–22. PubMed Abstract \| Publisher Full Text\n\nRitsch A, Scharnagl H, Eller P, et al.: Cholesteryl ester transfer protein and mortality in patients undergoing coronary angiography: the Ludwigshafen Risk and Cardiovascular Health study. Circulation. 2010; 121(3): 366–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobins SJ, Lyass A, Brocia RW, et al.: Plasma lipid transfer proteins and cardiovascular disease. The Framingham Heart Study. Atherosclerosis. 2013; 228(1): 230–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRidker PM, Paré G, Parker AN, et al.: Polymorphism of the CETP gene region, HDL cholesterol, and risk of future myocardial infarction: Genomewide analysis among 18 245 healthy women from the Women’s Genome Health Study. Circ Cardiovasc Genet. 2009; 2(1): 26–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoight BF, Peloso GM, Ortho-Melander M, et al.: Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study. Lancet. 2012; 380(9841): 572–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith GD, Ebrahim S: Mendelian randomization: prospects, potentials, and limitations. Int J Epidemiol. 2004; 33(1): 30–42. PubMed Abstract \| Publisher Full Text\n\nSleiman PM, Grant SF: Mendelian randomization in the era of genomewide association studies. Clin Chem. 2010; 56(5): 5723–8. PubMed Abstract \| Publisher Full Text\n\nBarsh GS, Copenhaver GP, Gibson G, et al.: Guidelines for genome-wide association studies. PLoS Genet. 2012; 8(7): e1002812. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSandhofer A, Tatarczyk T, Laimer M, et al.: The Taq1B-variant in the cholesteryl ester-transfer protein gene and the risk of metabolic syndrome. Obesity (Silver Spring). 2008; 16(4): 919–22. PubMed Abstract \| Publisher Full Text\n\nLeusink M, Onland-Moret NC, Asselbergs FW, et al.: Cholesteryl ester transfer protein polymorphisms, statin use and their impact on cholesterol levels and cardiovascular events. Clin Pharmacol Ther. 2014; 95(3): 314–20. PubMed Abstract \| Publisher Full Text\n\nPapp AC, Pinsonneault JK, Wang W, et al.: Cholesteryl Ester Transfer Protein (CETP) polymorphisms affect mRNA splicing, HDL levels, and sex-dependent cardiovascular risk. PLoS One. 2012; 7(3): e31930. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTanigawa H, Billheimer JT, Tohyama J, et al.: Expression of cholesteryl ester transfer protein in mice promotes macrophage reverse cholesterol transport. Circulation. 2007; 116(11): 1267–73. PubMed Abstract \| Publisher Full Text\n\nTchoua U, D’Souza W, Mukhamedova N, et al.: The effect of cholesteryl ester transfer protein overexpression and inhibition on reverse cholesterol transport. Cardiovasc Res. 2008; 77(4): 732–9. PubMed Abstract \| Publisher Full Text\n\nVillard EF, El Khoury P, Duchene E, et al.: Elevated CETP activity improves plasma cholesterol efflux capacity from human macrophages in women. Arterioscler Thromb Vasc Biol. 2013; 32(10): 2341–9. PubMed Abstract \| Publisher Full Text\n\nBorggreve SE, de Vries R, Dallinga-Thie GM, et al.: The ability of plasma to stimulate fibroblast cholesterol efflux is associated with the -629C-->A cholesteryl ester transfer protein promoter polymorphism: role of lecithin: cholesterol acyltransferase activity. Biochim Biophys Acta. 2008; 1781(1–2): 10–5. PubMed Abstract \| Publisher Full Text\n\nScharnag lH, Heuschneider C, Sailer S, et al.: Decreased cholesterol efflux capacity in patients with low cholesteryl ester transfer protein plasma levels. Eur J Clin Invest. 2014; 44(4): 395–401. PubMed Abstract \| Publisher Full Text\n\nYamashita S, Ishigami M, Arai T, et al.: Very high density lipoproteins induced by plasma cholesteryl ester transfer protein (CETP) have a potent antiatherogenic function. Ann N Y Acad Sci. 1995; 748: 606–8. PubMed Abstract \| Publisher Full Text\n\nShimoji E, Zhang B, Fan P, et al.: Inhibition of cholesteryl ester transfer protein increases serum apolipoprotein (apo) A-I levels by increasing the synthesis of apo A-I in rabbits. Atherosclerosis. 2004; 172(2): 247–57. PubMed Abstract \| Publisher Full Text\n\nDuverger N, Kruth H, Emmanuel F, et al.: Inhibition of atherosclerosis development in cholesterol-fed human apolipoprotein A-I-transgenic rabbits. Circulation. 1996; 15(4): 94: 713–7. PubMed Abstract \| Publisher Full Text\n\nBrousseau ME, Diffenderfer MR, Millar JS, et al.: Effects of cholesteryl ester transfer protein inhibition on high-density lipoprotein subspecies, apolipoprotein A-I metabolism, and fecal sterol excretion. Arterioscler Thromb Vasc Biol. 2005; 25(5): 1057–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOshima R, Ikeda T, Watanabe K, et al.: Probucol treatment attenuates the aortic atherosclerosis in Watanabe heritable hyperlipidemic rabbits. Atherosclerosis. 1998; 137(1): 13–22. PubMed Abstract \| Publisher Full Text\n\nYing H, Saku K, Harada R, et al.: Putative mechanisms of action of probucol on high-density lipoprotein apolipoprotein A-I and its isoproteins kinetics in rabbits. Biochim Biophys Acta. 1990; 1047(3): 247–54. PubMed Abstract \| Publisher Full Text\n\nBarter PJ, Rye KA, Beltangady MC, et al.: Relationship between atorvastatin dose and the harm caused by torcetrapib. J Lipid Res. 2012; 53(11): 2436–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarter PJ, Rye KA: Cholesteryl ester transfer protein inhibition as a strategy to reduce cardiovascular risk. J Lipid Res. 2012; 53(9): 1755–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDurrington PN: Cholesteryl ester transfer protein (CETP) inhibitors. Br J Cardiol. 2012; 19: 126–33. Publisher Full Text\n\nHewing B, Fisher EA: Rationale for cholesteryl ester transfer protein inhibitors. Curr Opin Lipidol. 2012; 23(4): 372–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShinkai H: Cholesteryl ester transfer protein inhibitors and their potential for the treatment of cardiovascular diseases. Vasc Health Risk Manag. 2012; 8: 323–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiyares MA, Davis K: Patient considerations and clinical impact of cholesteryl ester transfer protein inhibitors in the management of dydlipidaemia: focus on anacetrapib. Vasc Health Risk Manag. 2012; 8: 483–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\ndeGoma EM, Rader DJ: High-density lipoprotein particle number: a better measure to quantify high-density lipoprotein? J Am Coll Cardiol. 2012; 60(8): 517–20. PubMed Abstract \| Publisher Full Text\n\nKingwell BA, Chapman MJ, Kontush A, et al.: HDL-targeted therapies: progress, failures and future. Nat Rev Drug Discov. 2014; 13(6): 445–64. PubMed Abstract \| Publisher Full Text\n\nA study of evacetrapib in high risk vascular disease. Reference Source\n\nNicholls SJ, Brewer HB, Kastelein JJ, et al.: Effects of the CETP inhibitor evacetrapib administered as monotherapy or in combination with statins on HDL and LDL cholesterol: a randomized controlled trial. JAMA. 2011; 306(19): 2099–109. PubMed Abstract \| Publisher Full Text\n\nRandomized evaluation of the effects of anacetrapib through lipid-modification. Reference Source\n\nCannon CP, Shah S, Dansky HM, et al.: Safety of anacetrapib in patients with or at high risk for coronary heart disease. N Engl J Med. 2010; 363(25): 2406–15. PubMed Abstract \| Publisher Full Text\n\nKarlberg JPE, Speers MA: Reviewing clinical trials: A guide for ethics committees. 2010. Reference Source" }	[ { "id": "5462", "date": "17 Jul 2014", "name": "Gerald Watts", "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 exceptionally good review and balanced assessment of the status of CETP inhibitors and ASCVD from a world authority in the field. The article highlights important data that might have been overlooked when promulgating the clinical value of CETPIs and related trials.Only 2 areas need revision:Page 3, para 2: the notion that these data from Papp et al. convey is critical and the message needs an explicit sentence or two at end of paragraph. Page 4, Conclusion: the assertion concerning the ethics of the two Phase 3 clinical trials needs toning down. Perhaps rephrase to indicate that the value and sense of doing these trials is open to question, with attendant ethical implications, or softer wording to that effect.", "responses": [] }, { "id": "5614", "date": "29 Jul 2014", "name": "Cesare Sirtori", "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 Miller attempts to elucidate once and for all the present role of CETP and CETP antagonists in coronary prevention. The message is clear and well written. I do have a question on the structure of the paper however: while the background of CETP from the earlier papers by Havel is clear and well written, I question the idea of separating this from the very informative Table 1 which collates two separate pieces of information and is otherwise unclear. Also, as the current trend is to evaluate genomic analysis, in particular by Mendelian randomization, this could indicate that HDL is not a genetic risk component. This should be given special emphasis, since the Mendelian randomization analysis was focused on LIPC (just one determinant of HDL levels) whereas the author points out that SNPs of CETP are also probably involved. However he gives no explanation of the mechanism of these SNPs, which are apparently associated with raised HDL and also with reduced incidence of MI. In my view, the major focus of this paper should be on drugs, especially as drugs are still used to antagonize CETP in clinical trials. In the studies on torcetrapib it was noted (Nicholls et al, Circulation 2008) that individuals with the highest post-treatment HDL-cholesterol are apparently protected from coronary disease. Does this make sense? After all if HDL is traveling through the blood in very large amounts without going back to the artery this could provide some potential protection, although this should be evaluated with more rigor. A final note on the studies with torcetrapib is the incremental elevation of blood pressure. The rise of blood pressure was of such little entity (at most 5 mmHg) that just posting these data in a risk score (Framingham or other) allowed one to conclude that this influences risk minimally and does modify the enormous change in risk consequent to the HDL rise elicited by the CETP inhibitor (Sirtori, Mombelli, Clin. Chem, 2010). In regards to the rabbit studies: this started with the Okamoto paper in Nature (2000), which had one serious problem; the control group only had a final cholesterol of 129 mg/dl. Thus, in my view, the apparent advantage of giving JT -705 (dalcetrapib) makes no sense. A much better conducted study published by Huang (Huang et al, Clin Sci, 2002) used a similar protocol but had a cholesterol of 757 mg/dl in the control group. In this study, in spite of a marked rise of HDL-C, there was no arterial benefit. This, in my view, should have closed the story. Unfortunately it did not. The probucol data, on the other hand, are definitely of high significance. Probucol raises CETP and prevents arterial disease (a number of reports have recently come from Japanese investigators e.g. Kasai et al. Atherosclerosis. 2012). Most excitingly, it removes cholesterol deposits (xanthelasmas/xanthomas) thus indicating that increased CETP is beneficial in man. Another negative issue to be raised is that in the Brusseau paper (ref 40) there was no evidence of a reduced cholesterol pool/increased fecal steroid excretion following torcetrapib, indicating again that blocking CETP does not in any way improve cholesterol turnover. I tend to believe that the issue of apo A-I syntheis is not of major significance. The reduction of AI synthesis (probucol) and increased synthesis (torcetrapib) is therefore of little interest. In conclusion the author has certainly done a very good job but the paper would be improved by the inclusion of some older data, that in my view are more significant, vs data provided by the recent overviews or Mendelian randomizations.", "responses": [] } ]	1	https://f1000research.com/articles/3-124
https://f1000research.com/articles/2-20/v1	23 Jan 13	{ "type": "Method Article", "title": "A flexible user-interface for audiovisual presentation and interactive control in neurobehavioral experiments", "authors": [ "Christopher T Noto", "Suleman Mazhar", "James Gnadt", "Jagmeet S Kanwal", "Christopher T Noto", "Suleman Mazhar", "James Gnadt" ], "abstract": "A major problem facing behavioral neuroscientists is a lack of unified, vendor-distributed data acquisition systems that allow stimulus presentation and behavioral monitoring while recording neural activity. Numerous systems perform one of these tasks well independently, but to our knowledge, a useful package with a straightforward user interface does not exist. Here we describe the development of a flexible, script-based user interface that enables customization for real-time stimulus presentation, behavioral monitoring and data acquisition. The experimental design can also incorporate neural microstimulation paradigms. We used this interface to deliver multimodal, auditory and visual (images or video) stimuli to a nonhuman primate and acquire single-unit data. Our design is cost-effective and works well with commercially available hardware and software. Our design incorporates a script, providing high-level control of data acquisition via a sequencer running on a digital signal processor to enable behaviorally triggered control of the presentation of visual and auditory stimuli. Our experiments were conducted in combination with eye-tracking hardware. The script, however, is designed to be broadly useful to neuroscientists who may want to deliver stimuli of different modalities using any animal model.", "keywords": [ "Spike2", "Presentation", "multimodal stimuli", "data acquisition", "eye position", "neural recording", "inferior colliculus", "superior colliculus", "auditory" ], "content": "Introduction\n\nIn neurophysiological research, correlating neural signals driven by stimulus presentation and behavioral response needs to be completed within a limited time frame, generally less than 2 hours when conducted with non-human primates. This requires effective and efficient control of presentation of stimuli, acquisition of data, and monitoring of behavior for reward and task progression. Behavioral neuroscientists have to continuously struggle to both keep up with technological advances to accelerate data throughput and to customize stimulus delivery and data acquisition systems to do cutting-edge research. This adds to the burden of labor-intensive electrophysiological recordings from single or multiple neurons in awake-behaving animals, which nevertheless continues to be the most reliable and useful way to understand neural computations and function. Stimulus presentation paradigms may also need to be routinely modified to conform to the goals of an experiment. All of this has to be accomplished with the constraint of maintaining the experimental animal in a healthy condition until the experiment has run its course, which may take from weeks to months. Moreover, user requirements, dictated by the scientific data and state of knowledge, are a moving target that makes it difficult for for-profit vendors to meet all the needs of their customers. Laboratory heads are frequently faced with the task of either hiring a permanent programmer at the cost of tens of thousands of dollars in annual salary to create and maintain a new program, or abandoning a particular line of experiments that scientifically may be the right direction in which to proceed. Even the choice of hardware and software packages that laboratory personnel could interface with and manipulate easily largely depends upon the available expertise of those working in the laboratory and frequently shifts with the departure of key personnel.\n\nTo effectively meet our own needs for the study of gaze control in response to the presentation of audiovisual stimuli, we developed a user-interface that provides a template for others facing a similar challenge. Specifically, we describe an experimental design that uses a custom-written script for controlling communication between Presentation software (Neurobehavioral Systems, Inc., Albany CA) package and data acquisition hardware (Cambridge Electronic Design, Ltd., Cambridge, UK) along with vendor-provided Spike2 software. Each package runs independently on separate personal computers (Figure 1A).\n\nA. The system is divided into three levels: Data Acquisition and Behavioral Monitoring, Stimulus Presentation Control, and Stimulus Presentation. B. Typical flow of experiment based on performance of subjects. C. Logical flow chart for behavioral tasks used in the training and testing paradigms.\n\nRapid eye-movements, or saccades, channel important visual information into the association and prefrontal cortex where it is integrated with previous knowledge to take decisive action1–3. Much research effort has gone into validating the role of the superior colliculus (SC)4–7 and cortical areas, such as the frontal eye fields2 and lateral intraparietal areas1, in the control of eye movements during visual saccades. Less is known about the kinematic properties and control of auditory priming of saccades8,9 and virtually nothing is known about the modulation of visual saccades by contextual auditory information. Clearly, sensory recall of auditory objects as well as error-correction and decision mechanisms underlying memory-guided saccade initiation or head orientation need to be invoked9–11. Given past studies it is now possible to explore the location in the brain where these mechanisms might converge12,13.\n\nWe used our newly developed audiovisual presentation and control scripts to acquire new data on responses to auditory and visual stimuli in a reward-driven behavioral task that involved tracking eye movements in a nonhuman primate. Our goal was to facilitate the exploration of neurons that integrate multimodal sensory information from naturalistic stimuli to elicit adaptive behavior. As a first step, we trained monkeys to associate relatively novel sounds, including animal vocalizations, with images that were also considered novel for monkeys maintained in a captive environment. To begin to explore the pontine circuitry creating such associations as well as eye movements, we first used species-specific calls to identify complex auditory stimulus-driven neurons in the IC, and naturalistic images to identify visual and saccade-driven neurons in the SC. This narrowed our search space for finding audiovisual neurons, located potentially at the boundary region between the IC and the SC, and testing if reward modulated their activity. We focused on eye movements as the adaptive behavior since these can be accurately tracked, provide a rapid response, and are controlled by neural activity within the SC4–7.\n\n\nMaterials and methods\n\nWe used two software packages, Presentation (Neurobehavioral Systems, Inc., Albany CA) and Spike2 (Cambridge Electronic Design, Ltd) in conjunction with data acquisition hardware (Power1401plus, Cambridge Electronic Design, Ltd), to control stimulus presentation based on our subject’s behavior. This required communication between Presentation and Spike2 software via serial and parallel ports to either advance or terminate a subject’s task in real-time based on either correct or incorrect behavioral responses, respectively. Data acquisition at a relatively high sampling rate (0.1 ms resolution) by the Power1401) was performed concurrently with stimulus presentation and behavioral monitoring. Our design integrates hardware that is either routinely available in a neurophysiology laboratory or commonly available from vendors (Table 1 and Figure 1A). Presentation software is readily available from Neurobehavioral Systems for on-line download (http://www.neurobs.com). We chose Presentation because of its large, comprehensive scripting language and intuitive user interface, and because the software allowed a simple method to communicate via both serial and parallel ports of a personal computer running a Windows operating system. The software itself is easy to use and numerous example scripts make the language easy to learn. Spike2 and the Power1401plus are available for purchase from Cambridge Electronic Design (http://www.ced.co.uk/indexu.shtml). We chose Spike2 and the Power1401plus because of its extensive scripting capabilities, ease of use, and inherent control of data acquisition, independent sequencer control, and straightforward manipulation of both parallel and serial ports. Each software package runs independently on its own personal computer to avoid compromising processor resources. Under Presentation control, video is output by a Radeon 9250 video card on a 55” Visio flat panel HD TV and sound is output using a SoundBlaster audio card by a Bose speaker system.\n\nThree Rhesus monkeys (Macacca mullata; 2 males and 1 female) acquired from a research facility at Wake Forest University, were available during various stages of testing and data acquisition for the development of protocols described here. Compatible animals were housed in paired grooming/contact cages (~2.5 cubic meters), in a room with a light and dark cycle set by an automatic day/night timer (light from 6AM to 6PM daily) and with full view of colony mates in a large open room. Cages were continuously equipped with swings, mirrors, foraging devices and/or small toys. Daily care and medical maintenance of the animals, including a balanced diet of dry food formula, vegetables and fruit, were routinely provided. Environmental enrichment for the monkeys included playing of natural sounds, radio or TV and daily handling, mock grooming and socialization by laboratory personnel.\n\nAnimals were prepared for participation in experiment by performing two surgeries. For the first surgery, we implanted a head restraining device and one scleral eye coil. With the head secured in a stereotaxic device, a 5 cm midline incision was made in the scalp. Periosteum and muscle was retracted using blunt techniques and the calvarium scraped free of soft tissue. A 3 cm stainless steel bar, which fits a head restraining apparatus of the primate chair, was attached vertically to the calvarium using surgical stainless steel screws and a stainless steel recording chamber anchored to the skull using screws and a mound of acrylic bone cement14,15. The screws are mounted into small burr holes in the bone and buried in the bone acrylic along with the head post and electrical connectors. A scleral eye coil was implanted on one eye. Briefly, the conjunctiva was cut near the limbus and reflected to expose the sclera. A coil made of three turns of Teflon-insulated wire was sutured to the sclera using 6-0 Vicryl, and the conjunctiva was sutured back over the coil. The ends of the coil wire were led out of the orbit subdermally to the acrylic cap where they were attached to a small electrical connector. One week post-surgery, we began a daily task-specific training regimen. Once training proceeded to an acceptable level, generally within a few months, another aseptic surgery was performed to implant an eye coil on the second eye and one or two stainless steel recording chamber(s) were mounted into the head cap under stereotaxic guidance. The acrylic overlaying the appropriate portion of the skull was removed using dental burrs in a hand drill and a 15 mm craniotomy was made. Stainless steel recording cylinders were placed over the craniotomy and cemented into place with bone or dental acrylic. The sterile interior of each cylinder was secured with a threaded Teflon cap having a pressure-release vent.\n\nPost-surgical maintenance included prophylactic antibiotics for 7 days (Baytril, daily 2–5 mg/kg) and 2–5 days of narcotic analgesics (buprenorphine, 0.05–0.1 mg/kg BID) followed by 3–5 days of acetaminophen (5–10 mg/kg). Flunixin, a non-steroidal anti-inflamatory agent, was administered for 1 to 3 days (0.5–1 mg/kg). We also monitored body weight and food/water intake daily, and performed maintenance of the skin margin and cleaned the recording cylinders.\n\nDuring the behavioral training, the monkeys sat in the Plexiglass primate chair within a cube of magnetic field coils. To avoid recording of eye movements being confounded with head motion and to stabilize the head while electrodes are inserted in the brain, the head was restrained painlessly by clamping the head post to a device on the chair. To motivate the subjects to perform adequately, for five days per week they received their daily fluids as reward for proper behavior. When daily training or experiments are terminated prematurely, fluids are supplemented up to the normal daily level for that subject. Fluid intake was monitored and recorded daily. Additionally, pulpy fruit or vegetables were used to reward good behavior when returning the animal to the home cage.\n\nUsing standard behavioral shaping procedures, the animals were trained to fixate and to follow small visual or auditory stimuli by rewarding them with a drop of fruit juice from a gravity-fed “straw” for successfully completing each series of eye movements defined by the presentation of the stimuli. Training and experimental procedures were performed for no longer than 5 hours per day, usually for 1–3 hours. Animals exhibiting discomfort were readjusted within the chair or returned to their home cage. The daily manipulations for the animals did not produce pain or distress. The cooperative demeanor of the monkeys gives us reason to believe that they find the laboratory situation stimulating and the social interaction with the investigators satisfying.\n\nAll surgical and experimental procedures were performed in accordance with federal and institutional guidelines on the care and use of laboratory animals as part of protocols approved by the Georgetown animal care and use committee (protocol # 09–025).\n\nFigure 1B shows what is displayed on the screen and the actions of the subject in response to the presentation of a visual stimulus. Figure 1C is a logical flow diagram to show the various steps listed as 4 tasks in the experimental scheme. The tasks are described as follows:\n\n1. Association task:\n\nA sound is played and the associated target image is simultaneously presented at the center of screen. In our experiments, short (1 s) tone bursts and natural sounds (e.g. communication calls etc.) were presented at stimulus levels of ~80 dB SPL (decibels of sound pressure level).\n\n2. Left-right-association task:\n\nA sound is played and an associated target image is simultaneously presented centered at a horizontal location a user-selected distance from the center of the screen, either on the left or on right side (the decision to present left or right is decided randomly at run-time).\n\n3. Single distracter task:\n\n(a). A sound is played and at the same time a “green dot” is presented at the center with simultaneous presentation of the associated target image and a distracter image on either sides of the circle. The position of images is decided randomly on runtime.\n\n(b). The target image and distracter image are retained on screen and eye-focus is monitored.\n\n4. Multiple distracters task:\n\n(a). A sound is played and at the same time, a “green dot” is presented at the center with simultaneous presentation of an associated “target image” and multiple (user selected) distracter images at user-specified locations on the screen. The position of the images is deliberately kept fixed in this task.\n\n(b). The target image and distracter image are retained on the screen and eye-position is monitored.\n\nRunning the script described in Figure 2A provides a user-interface in Spike2 that begins a cascade of dialog boxes that request information relevant to the experiment (e.g. subject name) and the basic parameters needed to monitor the behavior of the subject (e.g. detection window size, reward duration). After supplying the basic information (Figure 2B), a list of experimental scenarios is presented to the user in order to select the condition a subject will face. We have programmed a number of saccade-related tasks that use one (or more) of eight audio stimuli to direct our subject’s behavior to learned associations of visual images. A check box arrangement indicates a combination of stimuli the user intends to use in the experiment. As well, a number of timing variables (‘Time to get on Target’, ‘Initial Fixation Time’, ‘Fixation Time for Reward’) are adjustable by the user. Clicking the ‘OK’ button collapses the association-training dialog box allowing the user to hit the ‘Run’ button to initiate the scenario or to select a different scenario. From this point forward, the parameters dialog, the experimental scenario dialog, and a quit option are always available as buttons to the user on the Spike2 program interface. Selecting another experimental scenario automatically names and saves the current data file while initiating data collection into a new file for the newly selected scenario. Clicking on the ‘quit’ button saves the current data file, terminates the presentation of the ongoing stimulus to the subject and ends execution of the Spike2 script.\n\nA. Flow chart showing the design of experiment control sequence shared between the four scripts. B. The initial interactive user-interface used to collect basic information about the experiment set-up. C. User-interface used to collect the initial parameters for behavioral monitoring of subjects during experiments.\n\nAlthough quite simple, our ‘Association Training Paradigm’ allowed us to illustrate the inner workings of 1) the “Spike2 control” script, 2) the “sequencer script”, and 3) the “presentation script” as they operate across all the current scenarios available to the user. Before going on, we should discuss what we expect from the program and subject, so we can better discuss the interweaving functions of these three scripts. Figure 1B shows the progression of stimuli if the subject succeeds across all phases of the trial or fails at any time in the trial. This task has three phases: 1) an initial black screen or timeout screen, 2) an initial fixation target, and 3) test stimulus presentation. During phase 1, behavior is not actively monitored. The duration of the timeout is set to 2 seconds in the presentation script. At the inception of phase 2, the sequencer acting through the script loaded to the Power1401 memory begins monitoring eye position. The subject must first acquire the target and maintain gaze on the target within a small “forgiveness” window for the user-defined epoch of time. Successful fixation of the target advances the scenario to phase 3 by a command issued first from the sequencer to the Spike2 ‘control’ script and then from the ‘control’ script to the presentation script. Failure results in a reset to the black screen and a brief timeout using the same flow from sequencer to presentation script. A response token is sent directly back from the presentation script to both the ‘control’ script and sequencer ensuring that all three scripts remain synchronized. Phase 3 consists of the presentation of our test stimuli, here the co-presentation of an image and sound. Successful fixation of the image within a forgiveness window, equal to the size of the image and for the user-defined time, initiated by a dialog box shown in Figure 3A, results in the delivery of a reward to the subject as commanded by the sequencer. Successful fixation or failure to look at the image commands a reset of the experimental process to the black screen for a 2 second refresh period.\n\nA. The list of proposed paradigms users may select from to start a session; currently available are Calibration and the Association Tasks, and include “walk” tasks that were not implemented in the present version. “Walks” are saccade tasks designed to use a single target that appears on a black screen in various locations, moving in patterns ascribed by the selection buttons and subsequent dialogs boxes that may be added by the user. They can be used to train the animal and record metrics of their eye movements. B. User-interface that allows for selection of the auditory-visual pairing used during the association training paradigm and the timing criteria necessary for successful completion of the task.\n\nOur stimulus delivery and data acquisition package consisted of four primary components that operate in conjunction with one another. A “sequencer” script written from within Spike2 is loaded to the Power1401 for real-time monitoring of eye position and saccades. Sequencer scripts (included in the “Sequencer Files” file below) are ultimately responsible for issuing commands that direct progression through a task and to reward the subject. Two scripts (included in the “Spike2 Control Scripts” file below) operate in the Spike2 software environment. The first script controls the interfaces into which a user inputs relevant control parameters. The second script provides the functional control between the scripts running on the Spike2 computer and script running on the Presentation computer. One script (included in the “Presentation Scripts” file below) runs in the Presentation software environment commanding the output of stimuli and communicating by connections of the parallel line (to the Power1401) and the serial line (to the computer running Spike2) a time-stamp indicating when the presentation script commands the presentation of a stimulus.\n\nThe sequencer script downloaded to the Power1401 module runs using an independent clock from the Spike2 computer, but communicates with it through a high-speed USB port. The sequencer script consists of two parts: 1) the initialization section and 2) the monitoring section. The initialization sections load the user-defined variables set while interacting with the dialog boxes created by the ‘interface’ script. In our example of the association-training scenario, the variables loaded are the edges of the forgiveness window, the three timing criteria (time to get on target, initial fixation time, and fixation time for reward), and reward pulse duration. The sequencer cannot act on these values directly so we convert them to sequencer-relevant values. The edges of the forgiveness window are converted from the user-defined values in degrees to digital-to-analog converter (DAC) values. The timing criteria and reward pulse duration are converted from milliseconds to sequencer steps per ms. The monitoring section is made up of the same number of sections as the scenarios or situations (here three), each with specific tasks. The first task checks that the subject acquires the fixation target after it is presented within the user-defined epoch of time. The state of fixation, success or failure, is sent to the Spike2 ‘control’ script. If the sequencer determines the subject has worked within the task bounds, the sequencer steps to the next phase of the monitoring section and waits for a confirmation that the scenario has advanced from the ‘control’ script. The second task checks that fixation is maintained on the target for the specified time. Once again, information about the state of fixation, either success or failure, is sent to the Spike2 ‘control’ script. If the sequencer determines the subject has worked within the task bounds, the sequencer steps to the final phase of the monitoring section and waits for a confirmation that the status has advanced from the ‘control’ script. The final phase operates exactly as the second phase except that if the subject complies, a reward pulse is sent from the Power1401 to a reward delivery system though one of the digital I/O ports. Regardless of whether the trial is deemed a success or failure, the sequencer returns to the initialization section and resets the variables to their initial user-defined state. The process loops with each trial.\n\nThe ‘control’ script runs in the background on the Spike2 computer and uses the first bit of the COM-1 port input to communicate with the Presentation computer and controls advancement through the Presentation script. Each task consists of a looping “do case” function with progress through the function determined by the fixation state passed from the sequencer. In our example of the association-training task, there exist three fixation failure situations and three successful fixation situations. The first failure scenario occurs when the screen is black and the subject has no target to fixate. The script simply calls for the presentation of the fixation target by issuing a command to the computer running Presentation. The second and third failure situations are similar and initiate a command for the presentation of the black screen to the subject. The first success case assumes that the subject’s gaze is directed toward the fixation target’s location when there is none present. In this situation, the script calls for the fixation target to be presented, just as in the first failure case. The second success situation calls for the Presentation script to display the test stimulus. The third success scenario initiates a reset of the screen to a blank (black) display by the Presentation script. Following each call to the Presentation computer, the ‘control’ script listens for a reply on bit 1 of the COM-1 port. Upon receipt of the response the “do case” state is returned to the sequencer to allow advancement through the monitoring sections of the script ensuring proper stepwise alignment of all three scripts throughout the task.\n\nThe Presentation script running on the Presentation computer acts as a slave to the Spike2 ‘control’ script receiving instructions and replying through the first bit of the COM port. This script has three primary sections: 1) the video monitor setup, 2) image and sound object creation, and 3) the experimental loop. The first section of this script requires the user to predefine the current display properties (resolution and color depth) including the height and width of the monitor, and distance of the subject to the screen. In this way, Presentation calibrates itself so that target and image positions may be stated in degrees and drawn at the appropriate size. The second section predefines all the potential objects that may be called during the experiment after selection via a dialog box (Figure 3B) and their association, if any. For example, our fixation target is a small green dot. We have created an object (e.g. named ‘greendot’) that holds all the relevant information about how presentation draws our fixation target (e.g. dot size, color of the background, etc.) when a call is made to the object. In the third section, the Experimental Loop monitors the COM-1 port for communication from the Spike2 computer. This loop is largely comprised of “if, then, else” statements. Each communication from the ‘control’ script is pre-defined so that when the ‘control’ script shunts words and terminators (e.g. ‘grendot\\n’) to the Presentation computer, the Experimental Loop recognizes the word (grendot) and terminator (\\n) and falls into the appropriate “if” statement. In the case of the ‘grendot\\n’ combination, the “if” statement calls for our object ‘greendot’ so that the fixation target is displayed on the monitor, and at the same time triggers a reply to the ‘control’ script on the COM-1 port and to the Power1401 on bit 8 of the parallel port (a 1 ms low-high-low transistor-transistor logic (TTL). The script then returns to the loop, listening for the next command from the Spike2 computer. In this way, each object may be called in any sequence as commanded by the Spike2 ‘control’ sequence. In the case of our example, the next word that the loop would receive would be ‘SndPICn\\n’. Similarly, the loop falls into the appropriate “if” statement, displays the test stimulus, replies to the Spike2 computer and Power1401, and returns to the loop.\n\nUp to five days a week, a two-hour neural recording period occurred between 10AM and 4PM to ensure overlap with veterinary staff hours. Animals were moved from their home cage to an adjacent room for neural recording sessions while seated comfortably in a primate chair. In the recording room, the animal’s head is fixed facing forward, in full view of a LCD (liquid crystal display) monitor set 48 inches in front of them with the center of the screen at the approximate height of the animals straight ahead gaze. Extracellular neuronal recordings were made using standard electrophysiological methods in behaving subjects using fine wire tungsten microelectrodes (31 gage, Microprobe, Inc.) mounted in a guide tube of stainless steel hypodermic tubing16. Transdural penetrations were made by a hydraulic microdrive (FHC, Inc.) advancing a tungsten electrode through the bore of a 21 gage hypodermic needle mounted in a micropositioner that attaches to the outside of the chronic recording cylinders on the animal’s head. Neuronal activity was recorded on the hard drive of a laboratory computer via a high impedance amplifier system (AMC Systems, Inc.).\n\nData analysis and recording was conducting using Spike2 software (Cambridge Electronic Design, Ltd.). Custom-written scripts were used to build raster plots and peristimulus – time histograms (PSTHs) for display of processed data for well-isolated single units whenever possible. Only sample data from single or few-unit activity are provided here to demonstrate feasibility for the purposes of this project, which was designed for development of experimental control procedures.\n\n\nResults and discussion\n\nAs proof of concept, we present here a number of behavioral and neural responses from various brain structures that are activated in response to naturalistic stimuli presented within our experimental set up. To reiterate, we were primarily concerned with capturing 4 basic types of neural responses: 1) visual, 2) auditory, 3) saccade, and 4) reward-driven. This analysis utilizes the timestamps placed in the data files by the presentation script’s 1 ms TTL pulse sent to the Power1401 during data acquisition. Neurons were recorded from the midbrain in the putative inferior and superior colliculi (IC and SC, respectively) of one of our nonhuman primate subjects. Figure 4 describes typical neural responses in the IC following the presentation of complex communication sounds or “calls” that contained acoustic features preferred by the neuron17–20. Of the ten neurons from which electrophysiological activity was recorded, all responded to at least one of the seven sounds presented. As an example, Figure 4 shows the response of two neurons from the same animal to the same three sounds. We found that each sound produced a distinct temporal response pattern. These patterns could range from no or transient increases in the overall firing rate (upper left panel) to intense phase-locked responses to acoustic features within a call (lower right panel).\n\nAmplitude envelops (top) and raster and PSTH (10 ms bins) plots superimposed on call spectrographs of three different call types (grunt, harmonic, noisy) to show the response of 2 neurons (neuron 1 is from a female and neuron 2 is from a male) in the monkey inferior colliculus. Each call presentation was repeated 40 times per histogram. Grey vertical dotted lines indicate sound onset. Note the response build-up to the third predominant amplitude modulation in the last call. Average first peak response latency to calls was 20.9 +/- 3.5 ms (n = 10). Responses with a potential for temporal facilitation are enclosed by ellipses, although response enhancement may also depend on the basic acoustic patterns within complex sounds or on amplitude tuning. Calls were downloaded from the following web site: http://www.soundboard.com/sb/Rhesus_Monkey_sounds.aspx.\n\nThe neuron shown in Figure 5A and 5B illustrates the characteristic visual activity one expects to find while recording from rostral-superficial layers of the SC21. Once gaze was directed to position the eyes within the receptive field of this neuron we observed steady, low-rate firing within ~20ms. In this example, the subject was required to make a saccade to capture a sound-associated image. After the fixation target was extinguished and the target image was presented in the peripheral field of vision, the neural response declined and resumed only when the eyes were positioned again on the target image. The neuron shown in Figure 5C and 5D fits the characteristics attributed to neurons of the intermediate layers of the SC22,23. Namely, a 60–80 ms build up in activity followed by a burst of spikes just prior to the initiation of direction-dependent saccades to our visual stimulus. Examination of the neural data collected during “spontaneous” eye movement behavior shows that this neuron preferred saccade vector (>20 degrees amplitude, 137 degrees angle), which is well off the axis of our stimulus.\n\nA. Summed histogram from multiple trials triggering visual stimulus-induced activity in a “fixation” neuron located at the rostral pole and superficial layers of the superior colliculus (SC). Electrophysiological responses (timestamps for spikes) were aligned to the time at which the subject acquired the fixation target to begin the trial. B. Saccade-triggered transient suppression of neural activity in a different neuron located within intermediate layers and caudal to the fixation-neuron in the same animal. Dashed vertical line is at time zero for stimulus presentation in “A” and for target fixation in “B”. The width of the grey bar indicates the neuron’s visual delay (>20ms). C. Raster plots (above) and binned profile of summed response (below) to compare neural activity during rightward (top panel) and leftward (bottom panel) saccades. Grey bins indicate the build-up phase while the black bins indicate the burst phase of the neuron. Dashed vertical line indicates saccade onset. D. Heat map of saccade-related neural activation. Black box enclosed by dashed lines indicates position of the target image relative to central gaze. Solid white arrow represents the vector for the preferred saccade as indicated by firing rate of the neuron.\n\nOf the total population of neurons studied in the SC by Jay and Sparks12, 79% showed saccade-related bursts prior to eye movements to either visual or auditory evoked target stimuli suggesting that saccades evoked by either stimulus share a common efferent pathway to generate the movement. Meredith et al.13 recorded 113 neurons in the SC (82/113 were auditory-visual neurons) of anesthetized cats during presentation of single and temporally overlapping sensory stimuli. Peak response in neural firing to multisensory signals occurred when stimuli were presented concurrently, with the second stimulus starting <100 ms from the first. Since then, research has shown that in the deep layers of the SC, most neurons respond to both visual and auditory stimuli; 99 of 121 SC neurons showed significant alteration in firing rates due to eye position12.\n\nApproximately 60% of neurons within the IC have been shown to respond to not only sound17, but to some extent visual- and saccade-related activity24,25. Inputs from the lateral nucleus of the IC and the nucleus of the brachium of the IC to the SC also exist26. This pathway may be responsible in part for the auditory activity observed in the deep layers of the SC27 and is one route via which auditory information can influence saccadic eye movements. The response of IC neurons to visual stimulus and during eye movements is much less robust than the activity observed following visual stimulation and during saccades in the SC. The use of natural stimuli is expected to boost the responses of IC neurons in an audiovisual recall task to reveal multi-sensory integration that can influence saccade-related activity. Figure 6 illustrates a neuron’s activity that is putatively considered reward-dependent28,29. The neuron was located rostral to the IC and deep to the region known to contain neurons controlling saccade-related activity in the SC. The neural activity was clearly phase-locked to the task, but was less obviously linked to auditory stimuli (Figure 6A), contrary to what one would ordinarily expect in IC neurons (compare with Figure 4). This activity was not strictly linked to visual stimuli, nor was it saccade-related in terms of SC activity. The activity of this neuron seemed to indicate an expectation of reward that builds up based on successfully meeting task-related milestones (Figure 6B). During the task and especially following the onset of the sound, very distinct differences existed in the firing pattern of this neuron compared to between the two conditions.\n\nA. Single experimental trial illustrating stimulus presentation and related neural activity. During these trials, the subject was rewarded for successful discrimination of a sound-associated (target) image and a distractor image. B. Spike density waveform averaged from 20 trials aligned on fixation target onset. All other behavioral and stimulus markers are centered at their average time of occurrence and grey boxes indicate the first standard deviation in event time.\n\nBoth auditory and visual activity in space is read out in the SC in a manner that is appropriate for generating accurate saccades to sounds and images, respectively, although visually evoked saccades have high velocity, greater precision and shorter reaction times than auditory evoked saccades4,11,30. These and many other findings clearly indicate an extensive auditory input to the SC. Briefly, visual information from the retina drives the development of and maintains a spatial representation of auditory space in the IC31–33. This has been demonstrated in owls33,34 and is believed to be true in mammals. We presume that retinal inputs pass through the optic tectum and the superficial layers of the superior colliculus (SCs) before converging on auditory areas in the IC (Figure 7). A pathway from the retina to SCs to IC is known to exist in mammals35. Over the long term, the convergence of visual and auditory signals reinforces an enduring spatial map in the IC. Recently, many neurons within the IC (the brachium of the IC, the external capsule of the IC, and the core of the IC), have been shown to respond to not only sound17, but to some extent visual- and saccade-related activity and in some cases responses are modifiable by reward24,25.\n\nIn many species, including humans, who rely predominantly on vision for their survival, auditory cues may trigger eye movements either for interaction with the environment or for communication with conspecifics. Many researchers point to the SC and IC as components of a multi-modal sensory integration system, where visual and auditory signals within the brain merge into a co-dependent representation of the world3–5,36,37. Neurophysiological and anatomical data support the idea that this linkage occurs only two or three synapses beyond the retina and auditory nerve. Signals sent out of the IC and SC are also fed-back onto their independent systems helping to modulate behavior (see Figure 7).\n\nThe superior (SC) and inferior colliculi (IC) receive direct visual (shades of green) and auditory (shades of red) projections, respectively and have reciprocal connections with each other. The IC also receives emotive inputs from the amygdala (AM) either directly43 or via reward circuitry in the ventral striatum44, and has reciprocal connections with the auditory cortex (AC) for cognitive processing48. Saccadic eye movements are controlled by outputs from the SC via local burst generators (BG) driving motor neurons innervating extraocular muscles (EOM). The SC receives information from the visual cortex and premotor neurons in the frontal cortex, particularly the frontal eye fields. AC = Auditory Cortex; CN = cochlear nucleus (VIIIn); MN = motor neurons (nuclei of cranial nerves III, IV and VI); PM = Pre-motor Cortex; VC = Visual Cortex.\n\nIn summary, collecting behavioral and neural data using our suite of scripts and hardware together with subsequent analysis yielded new insights, providing strong evidence for the advantage of using a novel and customized paradigm. Our scripted user-interface demonstrated that pairing auditory and visual stimuli caused modest changes in activity throughout the trial period in a receptive neuron located deep within the SC. This was in contrast to the response of the same neuron presented with the same stimuli when the animal listened to them passively. The SC appears to be the site where sensory signals encoded in different frames of reference converge, and are translated into a common coordinate system commanding movement execution11 (e.g. retinotopic-centered commands to resolve motor error). Integration of auditory and visual information also appears to occur at this site. A major cortico-collicular auditory projection suggests that the cortex may direct this integration via the IC, particularly during the learning phase38,39. After that, subcortical circuits may function autonomously for computing a reaction.\n\n\nConclusions\n\nIn conclusion, we have developed a simple and relatively straightforward user-interface that directs and monitors subject behavior as well as acquires data. This particular set-up and the customized paradigms used in this experiment may be impossible for vendors of commercial stimulus presentation and data acquisition software and hardware to develop for the general neuroscience community due to the specific needs of each research laboratory. Our experimental design and custom scripts, however, are flexible to meet virtually all experimental control and data acquisition needs of those interested in conducting behaviorally controlled, response-based experiments. We have used a modified design to run psychophysics experiments on human subjects and these can be combined with dense array EEG recordings in response to the presentation of auditory and visual stimuli40. In essence, our template can be used to build any type of subject-interactive experiments. There is high potential for applying our pragmatic design to control neurobehavioral experiments using readily available hardware and software. Our studies, using earlier methodologies, showed that arousal has a role in bottom-up modulation of thalamic activity in the control of eye-movements41,42. Our new methodology allowed us to discover the location of audiovisual neurons at which reward-based, and possibly anxiety-driven, influences may converge to modulate behavior43–47. Studying these circuits in intact, normal animals is important to decipher the interplay of excitation and inhibition between different neural circuits for dynamic control of eye movements and gaze control.", "appendix": "Author contributions\n\n\n\nCTN: Participated in experimental design, surgical procedures, writing scripts, data acquisition and analysis, and writing the manuscript.\n\nSM: Participated in experimental design, assisted with surgical procedures, writing scripts, and data acquisition.\n\nJG: Designed original experimental set up, participated in scientific discussions, animal acquisition, care and surgeries.\n\nJSK: Conceptualized and participated in experimental design, data analysis and writing the manuscript.\n\nAll authors read and agreed to the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests have been disclosed.\n\n\nGrant information\n\nWork supported in part by grant EY015870 to JSK from the National Eye Institute (NEI). We also thank the Biomedical Graduate Research Organization (BGRO) of Georgetown University for financial support to JSK during the later phase of this project. The content is sole responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.\n\n\nAcknowledgements\n\nMr. Dolphus Truss was especially helpful in assisting with animal care issues.\n\n\nReferences\n\nAndersen RA, Mountcastle VB: The influence of the angle of gaze upon the excitability of the light-sensitive neurons of the posterior parietal cortex. J Neurosci. 1983; 3(3): 532–548. PubMed Abstract\n\nBichot NP, Schall JD, Thompson KG, et al.: Visual feature selectivity in frontal eye fields induced by experience in mature macaques. Nature. 1996; 381(6584): 697–699. PubMed Abstract \| Publisher Full Text\n\nMiller EK, Cohen JD: An integrative theory of prefrontal cortex function. Annu Rev Neurosci. 2001; 24: 167–202. PubMed Abstract \| Publisher Full Text\n\nSparks DL: Response properties of eye movement-related neurons in the monkey superior colliculus. Brain Res. 1975; 90(1): 147–152. PubMed Abstract \| Publisher Full Text\n\nGuthrie BL, Porter JD, Sparks DL, et al.: Corollary discharge provides accurate eye position information to the oculomotor system. Science. 1983; 221(4616): 1193–1195. PubMed Abstract \| Publisher Full Text\n\nSparks DL: The neural encoding of the location of targets for saccadic eye movements. J Exp Biol. 1989; 146: 195–207. PubMed Abstract\n\nSparks DL, Nelson JS: Sensory and motor maps in the mammalian superior colliculus. TINS. 1987; 10(8): 312–317.Publisher Full Text\n\nLewis JW, Wightman FL, Brefczynski JA, et al.: Human brain regions involved in recognizing environmental sounds. Cereb Cortex. 2004; 14(9): 1008–1021. PubMed Abstract \| Publisher Full Text\n\nZella JC, Brugge JF, Schnupp JW, et al.: Passive eye displacement alters auditory spatial receptive fields of cat superior colliculus neurons. Nat Neurosci. 2001; 4(12): 1167–1169. PubMed Abstract \| Publisher Full Text\n\nValentine DE, Moss CF: Spatially selective auditory responses in the superior colliculus of the echolocating bat. J Neurosci. 1997; 17(5): 1720–1733. PubMed Abstract\n\nBergeron A, Matsuo S, Guitton D, et al.: Superior colliculus encodes distance to target, not saccade amplitude, in multi-step gaze shifts. Nat Neurosci. 2003; 6(4): 404–413. PubMed Abstract \| Publisher Full Text\n\nJay MF, Sparks DL: Sensorimotor integration in the primate superior colliculus. II. Coordinates of auditory signals. J Neurophysiol. 1987; 57(1): 35–55. PubMed Abstract\n\nMeredith MA, Nemitz JW, Stein BE, et al.: Determinants of multisensory integration in superior colliculus neurons. I. Temporal factors. J Neurosci. 1987; 7(10): 3215–3229. PubMed Abstract\n\nRobinson DA: A method of measuring eye movement using a scleral search coil in a magnetic field. IEEE Trans Biomed Eng. 1963; 10: 137–145. PubMed Abstract\n\nRamcharan EJ, Gnadt JW, Sherman SM, et al.: Single-unit recording in the lateral geniculate nucleus of the awake behaving monkey. Methods. 2003; 30(2): 142–151. PubMed Abstract \| Publisher Full Text\n\nNoto CT, Gnadt JW: Saccade trajectories evoked by sequential and colliding stimulation of the monkey superior colliculus. Brain Res. 2009; 1295: 99–118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVersnel H, Zwiers MP, van Opstal AJ, et al.: Spectrotemporal response properties of inferior colliculus neurons in alert monkey. J Neurosci. 2009; 29(31): 9725–9739. PubMed Abstract \| Publisher Full Text\n\nPortfors CV: Combination sensitivity and processing of communicaiton calls in the inferior colliculus of the moustahced bat, Pteronotus parnellii. An Acad Bras Cienc. 2004; 76(2): 253–257. PubMed Abstract \| Publisher Full Text\n\nSuta D, Kvasnak E, Popelar J, et al.: Representation of species-specific vocalizations in the inferior colliculus of the guinea pig. J Neurophysiol. 2003; 90(6): 3794–3808. PubMed Abstract \| Publisher Full Text\n\nKlug A, Bauer EE, Hanson JT, et al.: Response selectivity for species-specific calls in the inferior colliculus of Mexican free-tailed bats is generated by inhibition. J Neurophysiol. 2002; 88(4): 1941–1954. PubMed Abstract\n\nGoldberg ME, Wurtz RH: Activity of superior colliculus in behaving monkey. I. Visual receptive fields of single neurons. J Neurophysiol. 1972; 35(4): 542–559. PubMed Abstract\n\nMohler CW, Wurtz RH: Organization of monkey superior colliculus: intermediate layer cells discharging before eye movements. J Neurophysiol. 1976; 39(4): 722–744. PubMed Abstract\n\nWurtz RH, Goldberg ME: Superior colliculus cell responses related to eye movements in awake monkeys. Science. 1971; 171(3966): 82–84. PubMed Abstract \| Publisher Full Text\n\nPorter KK, Metzger RR, Groh JM, et al.: Visual- and saccade-related signals in the primate inferior colliculus. Proc Natl Acad Sci U S A. 2007; 104(45): 17855–17860. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPorter KK, Metzger RR, Groh JM, et al.: Representation of eye position in primate inferior colliculus. J Neurophysiol. 2006; 95(3): 1826–1842. PubMed Abstract \| Publisher Full Text\n\nKing AJ, Jiang ZD, Moore DR, et al.: Auditory brainstem projections to the ferret superior colliculus: anatomical contribution to the neural coding of sound azimuth. J Comp Neurol. 1998; 390(3): 342–365. PubMed Abstract \| Publisher Full Text\n\nSparks DL: Neural cartography: sensory and motor maps in the superior colliculus. Brain Behav Evol. 1988; 31(1): 49–56. PubMed Abstract \| Publisher Full Text\n\nIkeda T, Hikosaka O: Reward-dependent gain and bias of visual responses in primate superior colliculus. Neuron. 2003; 39(4): 693–700. PubMed Abstract \| Publisher Full Text\n\nIkeda T, Hikosaka O: Positive and negative modulation of motor response in primate superior colliculus by reward expectation. J Neurophysiol. 2007; 98(6): 3163–3170. PubMed Abstract \| Publisher Full Text\n\nSparks DL: Conceptual issues related to the role of the superior colliculus in the control of gaze. Curr Opin Neurobiol. 1999; 9(6): 698–707. PubMed Abstract \| Publisher Full Text\n\nKnudsen EI: Early auditory experience aligns the auditory map of space in the optic tectum of the barn owl. Science. 1983; 222(4626): 939–942. PubMed Abstract \| Publisher Full Text\n\nKnudsen EI, Knudsen PF: Visuomotor adaptation to displacing prisms by adult and baby barn owls. J Neurosci. 1989; 9(9): 3297–3305. PubMed Abstract\n\nKnudsen EI, Knudsen PF: Vision calibrates sound localization in developing barn owls. J Neurosci. 1989; 9(9): 3306–3313. PubMed Abstract\n\nMiller GL, Knudsen EI: Early auditory experience induces frequency-specific, adaptive plasticity in the forebrain gaze fields of the barn owl. J Neurophysiol. 2001; 85(5): 2184–2194. PubMed Abstract\n\nDoubell TP, Baron J, Skaliora I, et al.: Topographical projection from the superior colliculus to the nucleus of the brachium of the inferior colliculus in the ferret: convergence of visual and auditory information. Eur J Neurosci. 2000; 12(12): 4290–4308. PubMed Abstract \| Publisher Full Text\n\nSparks DL, Mays LE, Porter JD, et al.: Eye movements induced by pontine stimulation: interaction with visually triggered saccades. J Neurophysiol. 1987; 58(2): 300–318. PubMed Abstract\n\nGlimcher PW, Sparks DL: Effects of low-frequency stimulation of the superior colliculus on spontaneous and visually guided saccades. J Neurophysiol. 1993; 69(3): 953–964. PubMed Abstract\n\nYan J, Suga N: Corticofugal modulation of time-domain processing of biosonar information in bats. Science. 1996; 273(5278): 1100–1103. PubMed Abstract \| Publisher Full Text\n\nPerales M, Winer JA, Prieto JJ, et al.: Focal projections of cat auditory cortex to the pontine nuclei. J Comp Neurol. 2006; 497(6): 959–980. PubMed Abstract \| Publisher Full Text\n\nRana N, Medvedev AV, Noto CT, et al.: Brain activation in response to sounds signaling distress: Time-frequency analysis of EEG activity. Soc Neurosci. 2011; Program # 171.01, Washington DC, USA. Reference Source\n\nNoto CT, Gnadt JW, Kanwal JS, et al.: Neurometric and psychometric performance of Rhesus monkeys during a contrast discrimination task. Soc Neurosci. 2009; Program # 651.4, Chicago, IL, USA. Reference Source\n\nNoto CT, Gnadt JW, Beex LM, et al.: Parabrachial brainstem activation enhances contrast detection in macaque monkeys: Psychometric and neurometric analyses. Soc Neurosci. 2010; Program # 674.3, San DIego, CA. Reference Source\n\nMarsh RA, Fuzessery ZM, Grose CD, et al.: Projection to the inferior colliculus from the basal nucleus of the amygdala. J Neurosci. 2002; 22(23): 10449–10460. PubMed Abstract\n\nFudge JL, Kunishio K, Walsh P, et al.: Amygdaloid projections to ventromedial striatal subterritories in the primate. Neuroscience. 2002; 110(2): 257–275. PubMed Abstract \| Publisher Full Text\n\nEichenberger GC, Ribeiro SJ, Osaki MY, et al.: Neuroanatomical and psychopharmacological evidence for interaction between opioid and GABAergic neural pathways in the modulation of fear and defense elicited by electrical and chemical stimulation of the deep layers of the superior colliculus and dorsal periaqueductal gray matter. Neuropharmacology. 2002; 42(1): 48–59. PubMed Abstract \| Publisher Full Text\n\nNobre MJ, Brandao ML: Modulation of auditory-evoked potentials recorded in the inferior colliculus by GABAergic mechanisms in the basolateral and central nuclei of the amygdala in high- and low-anxiety rats. Brain Res. 2011; 1421: 20–29. PubMed Abstract \| Publisher Full Text\n\nDagnino-Subiabre A, Perez MA, Terreros G, et al.: Corticosterone treatment impairs auditory fear learning and the dendritic morphology of the rat inferior colliculus. Hear Res. 2012; 294(1–2): 104–113. PubMed Abstract \| Publisher Full Text\n\nYukie M: Connections between the amygdala and auditory cortical areas in the macaque monkey. Neurosci Res. 2002; 42(3): 219–229. PubMed Abstract \| Publisher Full Text" }	[ { "id": "856", "date": "21 Mar 2013", "name": "Bruce Cumming", "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 authors describe their system for presenting audio-visual stimuli and making neural recordings, by combining two packages, 'Presentation' and 'Spike2'. It sounds like a very reasonable arrangement for their purposes. But they do not make a case that this system achieves anything special methodologically. People have been doing experiments similar to those they describe for many (20+) years with a variety of different systems. It seems that the advance here is that this system is easy to use/set up. This is very likely to be true provided users do substantially similar experiments. But that’s true for almost any system. How easy would it be to do something else? Without some way of demonstrating in some objective way what the range of the system is, it is impossible to say if there is real methodological contribution here. Smaller questionsAn important problem is how precisely video events are synchronized with the neurophysiology. The moment at which a display computer requests an image change and the moment at which the first pixels of the new image are actually displayed on their LCD are two different things, and may well not even have a fixed delay (depending on the details of both the display and the rendering). How is this achieved? What is the delay? How variable is it? These crucial parameters are not reported.What is the total loop delay from detecting some event in the A/D stream (eye movement, Spike) and the change in some output (electrical stimulus, image refresh with a new image)? If detecting these events depends upon the Spike2 control script, then delays can be quite long. Implementing them in the Sequencer is much harder. This loop delay potentially places fundamental limits on the range of applications that might be possible (whether gazed contingent displays are possible, or performing cancellation tests with antidromic stimulation). Since the paper objective is to describe a software/hardware system, the details about surgery and training, and most of the results, seem irrelevant. This space would be better used describing measures of the system performance. In principle one might do something similar combining other separate systems e.g. any other commercial electrophysiology system and psychophysics toolbox for the display. What are the merits of the different possibilities? Without comparing the available options, this description is of limited use.", "responses": [ { "c_id": "530", "date": "22 Aug 2013", "name": "Jagmeet S. Kanwal", "role": "Author Response", "response": "We thank the referee for taking out the time to review our manuscript and have addressed the issues raised in the revision. Below we respond to each comment. We offer those so inclined to use these two packages a clear, working set of scripts as an example that can be used as a springboard in the development of their own experiments. In this way, our paper makes a significant methodological contribution, saving others precious time to focus on their experimental design. We were motivated by the F1000Research journal’s ability to provide a mechanism for archiving stimuli, scripts and raw data that the research community at large can readily access.Smaller questions:We agree that LCD displays include an inherent error in the timing of the display of images because of their refresh rates (at 60Hz the error could be up to ~16ms per frame). The error can be exacerbated by slow video rendering by the video card, which nowadays is less of an issue than in the past. The delay in our system was ~16ms, but the exact error (undetermined) is likely much less. We now include information on these issues in the revised version of the manuscript.The analog-to-digital converter (ADC) cycles through the analog signals at 1MHz (providing a read rate of 1 µs for the behavioral monitoring script loaded to the 1401’s sequencer); one line of code is acted upon with each tick of the clock. The sequencer has direct access to the ADC signals. The monitoring of events in this way is dependent on the length of the sequencer script loop. In our example, we sampled the position of the eye every 9 µs while in the loop. This means reading even a couple of hundred lines of code takes an order of magnitude less time for loop delays, allowing implementation of cancellation tests with antidromic stimulation, if needed. As indicated earlier, the delays of concern originate more with display issues for visual stimulation using naturalistic stimuli (simpler stimuli can be presented by other means), but these are still much smaller than perceptual delays, which are on the order of a couple of hundred milliseconds. For perturbing subconscious perception, either display delays could be accounted for in a stimulation paradigm or an LCD display may be substituted with a motorized slide projector. We describe relevant details regarding system performance as well as offer the results obtained in our study as a proof of concept, highlighting a potential scientific advance that may not be possible with a more cumbersome system. We agree that there are many options available for conducting neurophysiological studies and sometimes making the best choice can be difficult. A comparison between specific options would be helpful, but is somewhat arbitrary in the absence of knowledge of all available equipment and experimental goals. Therefore, we simply provide one example for a specific set of experiments and leave it to the reader to make the necessary comparisons given their objectives and available equipment. For this reason, we also believe it is important to illustrate the usage of the proposed setup in generating new findings and describe them adequately as well as provide animal protocols, as indicated by the editors/publishers. To that end, this manuscript accomplishes a clear and specific methodological goal." } ] }, { "id": "863", "date": "25 Mar 2013", "name": "Vincent Ferrera", "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 describes an interesting but fairly standard set-up for neurophysiology data acquisition in behaving subjects. It would be nice to have more details about how timing and synchronization were verified, and also what are the special capabilities of the hardware that justifies their selection? Nevertheless, it is good to know what other people are using and this is therefore a useful contribution.Minor comments The Power 1401 is rather expensive. What special capabilities justify this cost? Introduction: “which nevertheless continues to be the most reliable and useful way to understand neural computations and function.” This statement is debatable. The communication between the spike2 and presentation computers appears critical. More details would be welcome. What exactly is conveyed by the 1-bit com port? What is the signal timing diagram? How is this bit manipulated? A critical issue regarding the use of LCD displays for neurophysiology experiments is stimulus timing. How did the experimenters determine the relationship between the video output of the presentation computer and the actual appearance of stimuli on the display? Not clear about the hardware for neural signals. Was neural data stored on a third computer? Or on the spike2 computer? How many channels, what sampling rate, etc? How were spikes detected – window discriminator, waveform analysis?", "responses": [ { "c_id": "531", "date": "22 Aug 2013", "name": "Jagmeet S. Kanwal", "role": "Author Response", "response": "Thank you for taking the time to review our manuscript. Below we provide our detailed response to specific comments per changes made in the updated version of the manuscript. We generally agree with the reviewer’s point. Our main objective was to document an example for the benefit of others, given that the F1000Research journal provides a mechanism for archiving stimuli, scripts and raw data that the research community at large can readily access. We provide additional details in the revised version of the manuscript as suggested by the reviewer. Specifically, we include information relating to our data collection sampling parameters, serial line communication timing, analog-to-digital converter (ADC) sampling rates and clock speeds for sequencer monitoring of behavior.Minor commentsThe technical specifications can be found at: http://www.ced.co.uk/pru.shtml. We now cite this resource in the revised version. We specifically choose the Power 1401 because the hardware is capable of converting analog signals at high (1MHz) rates while simultaneously reading the ADC channels to allow the user to monitor subject behavior. Thus, the behavioral monitoring script loaded into the 1401’s sequencer at a 1 µs tick rate; i.e., one line of code was read with each tick of the clock. Combining these two capabilities gave us practically instantaneous monitoring of our subject and the ability to respond to their behavior with sub-millisecond precision. In the examples presented, we simply rewarded the animal for maintaining its eye position within a defined spatiotemporal window. One could extend this monitoring by defining a threshold on eye velocity, using that threshold to output a TTL pulse to trigger an external stimulator with similar precision.This statement now reads: “which nevertheless continues to be one of the most reliable and useful ways to understand neural computations and function.”Bit-1 of the serial line, com port, conveys the hexadecimal representation of the words (descriptors and terminators) used to call images and sounds and response tokens between acquisition and presentation computers, respectively. This bit was opened, written to, and closed by the respective portions of the scripts running on the acquisition or presentation computers. The signal passed between computers in less than a millisecond. We are unclear on which “signal timing diagram” is referred to by the reviewer since we do not use this term in the manuscript.We agree that LCD displays include an inherent error in the timing of the display of images because of their refresh rates (at 60Hz, the error could be up to ~16ms per frame). The error can be exacerbated by slow video rendering by the video card, which nowadays is less of an issue than in the past. We did not determine the exact error of our system. However, the response time of the display was 8 ms, which is half the refresh rate meaning that the system carries a maximum error of approximately 16 ms to refresh the entire image. For computing absolute perceptual or behavioral response delays, the refresh rate can be subtracted from the timing of the behavioral response, e.g., in our case eye-movements, though we were not interested in these particular parameters, only in using eye movements to control stimulus presentation and reward delivery.The neural signal passed through a filter and amplifier before undergoing analog-to-digital conversion by the 1401. We stored the data on the same computer running the acquisition script. We typically collect one channel of raw neural signal at either 25kHz or 50kHz, four channels corresponding to horizontal and vertical eye position at 1kHz, one auditory channel at 25kHz, and one channel of timestamps at 10kHz, generated on-the-fly during acquisition of data using an adjustable threshold set on the channel collecting the neural data on spike timestamps. Digitizing the raw neural signal allowed for post-hoc analysis using the Spike2 software that provides software window discriminators and level detectors as well as various forms of waveform analysis including template matching and spike sorting using PCA algorithms." } ] }, { "id": "963", "date": "23 May 2013", "name": "Farrel 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\nThis title is appropriate. The content of this article clearly describes the authors’ interface, components, and connections and the behaviour of this system. The authors support their assertion that they have developed a flexible and useful system for controlling and recording behaviour experiments. The authors show enough appropriate examples to support their assertions.", "responses": [] } ]	1	https://f1000research.com/articles/2-20
https://f1000research.com/articles/3-20/v1	21 Jan 14	{ "type": "Research Article", "title": "Policies to increase the social value of science and the scientist satisfaction. An exploratory survey among Harvard bioscientists.", "authors": [ "Andrea Ballabeni", "Andrea Boggio", "David Hemenway", "Andrea Boggio", "David Hemenway" ], "abstract": "Basic research in the biomedical field generates both knowledge that has a value per se regardless of its possible practical outcome and that has the potential to produce more practical benefits. Policies can increase the benefit potential to society of basic biomedical research by offering various kinds of incentives to basic researchers. In this paper we argue that soft incentives or “nudges” are particularly promising. However, to be well designed, these incentives must take into account the motivations, goals and views of the basic scientists. In the paper we present the results of an investigation that involved more than 300 scientists at Harvard Medical School and affiliated institutes. The study shows that basic researchers’ support for soft incentives is such that the transformative value of fundamental investigations can be increased without affecting the spirit of the basic research and scientists’ work satisfaction. After discussing the findings, we suggest a few examples of nudges and discuss one in more detail.", "keywords": [ "Basic or fundamental research—generally defined as untargeted research seeking to expand knowledge—is a key component of innovation. While it generates knowledge that has a value per se regardless of its possible practical outcome", "it also delivers knowledge that has the potential to produce more practical benefits1. Basic biomedical research in particular is crucial in addressing the challenges we face in our highly interconnected planet in which communicable diseases spread quickly and in which non-communicable diseases cause the premature death of many individuals2." ], "content": "Introduction\n\nBasic or fundamental research—generally defined as untargeted research seeking to expand knowledge—is a key component of innovation. While it generates knowledge that has a value per se regardless of its possible practical outcome, it also delivers knowledge that has the potential to produce more practical benefits1. Basic biomedical research in particular is crucial in addressing the challenges we face in our highly interconnected planet in which communicable diseases spread quickly and in which non-communicable diseases cause the premature death of many individuals2.\n\nHistorically, a wide range of basic biomedical research projects have contributed to the advancement of knowledge, from research solely inspired by the researcher’s curiosity to projects driven by a vision of how knowledge generated by research could be used as the basis for applied research. All research along this continuum is considered “basic” because it serves as the foundation for further research that may lead to applications. Scientific knowledge is produced by the coming together of all kinds of research streams and ideas. Abraham Flexner captured this aspect of science in the image of the Mississippi river, which “begins in a tiny rivulet in the distant forest. Gradually other streams swell its volume. And the roaring river that bursts the dikes is formed from countless sources”3. Although it often takes decades to develop, the applied outputs of knowledge advancement (e.g. drugs) have at their roots countless basic investigations.\n\nGiven its importance, complexity and breadth, basic research has been primarily funded by public money. This was particularly true in the decades that followed World War II and during which basic research went through a “golden age,” being conducted primarily in research universities and paid for with public money4. Sadly, public expenditure for research has decreased since then and nowadays fundamental sciences are for the most part underfunded. Basic biomedical research currently receives less support than it received only a few years ago. For instance, in the United States, National Institutes of Health (NIH) funding has been nearly stationary since 2003 in the face of rapid expansion of research activity in existing biomedical fields and the emergence of new ones1. One explanation of the low support for fundamental sciences is our cognitive bias in favor of immediate rewards. As our brains are structured in a way that leads us to unduly favor immediate rewards over future benefits5, we tend to underestimate the importance of human activities and initiatives with benefits that lie in the future6. Furthermore, basic researchers are said to owe a moral duty to extract maximum transformative value (the potential to translate into novel and fruitful applied research) whenever their research is publicly funded7,8. In an effort to maximize transformative value of research, funding agencies like the NIH already take in some consideration the health benefit potential or (social) “significance” of a research proposal when assigning resources.\n\nAssessing the societal impact of fundamental sciences is however more easily said than done. The topic is debated in the scientific and political communities9. Some policies already try to achieve this goal. Those based on strong financial incentives, such as the Bayh-Dole Act in the United States are already being used but they raise concerns as to whether monetary incentives distract basic scientists from focusing on fundamental questions10,11. New strategies to successfully maximize the transformative value of basic research without compromising the nature of fundamental inquiries are certainly needed. Softer incentives, which are called by behavioral economists “nudges”12, seem particularly promising as they have been successfully used in public health but not yet in basic research. If properly designed for basic research, nudges would slightly (and sometimes unperceivably) reorient some scientists in a certain direction without imposing rules or decreasing work satisfaction. However, these “nudges” can be designed well for basic research only if we have a good grasp of what motivates the basic scientists, what their values are and the intellectual frameworks in which they operate so that the proper soft incentives can be tailored around the particular characteristics of basic scientists.\n\nIn an effort to improve our understanding of basic scientists’ motivations, we designed a study to collect data from basic scientists at Harvard Medical School and affiliated institutions. We designed a survey that was filled out by more than 300 scientists. In the next sections, we present the results of this study and a discussion on how these findings can be used to increase the transformative value of basic biomedical research without decreasing the “basic” nature of these investigations and the motivations and freedom of the scientists. Finally, we suggest a few examples of nudges and discuss one in more detail.\n\n\nResults\n\nThe survey was designed as an online questionnaire comprised of 17 questions (Q1–Q17). On average, fewer than 2% of the respondents, with a range from 0% (Q2 and Q3) to 6.3% (Q15) skipped any of the 17 questions. Answers could be provided through multiple choices or, alternatively, textboxes for alphanumerical entries (see Methods section for additional details). 304 scientists took the survey. The first four questions of the questionnaire (Q1–Q4) gathered data on the sample characteristics. The first question (Q1) (all questions hereinafter will be referred to as Q#) aimed at identifying the respondent’s academic position: 39.9% declared themselves to be principal investigators, 34.7% to be post-docs, 10.6% to be PhD students and 14.9% to belong to other categories (including “research assistants” and “research technicians”) (Figure S1). Q2 focused on gender. The sample’s gender distribution turned out to be 42.1% females and 57.9% males (Figure S2). Q3 asked respondents to quantify the amount of their research time allocated to research that they consider to be “basic”. On average, respondents reported spending 76.3% of their research time on basic research (Figure S3), with only 3.6% of respondents stating that they were not involved (0%) in basic research. In order to test respondents’ commitment to basic research in general, Q4 asked them if they agreed/disagreed with the following statement: “Despite the current economic situation, public funding for basic biological/biomedical research should be increased”. 92.4% of the respondents agreed with the previous sentence while only 7.6% disagreed (Figure S4). Overall, these results show that our purposive sample was well balanced with regard to academic position and gender and that the surveyed scientists were significantly involved in basic investigations and (for the great majority) supportive of increased public funding for basic biological/biomedical research.\n\nThe way scientists conceptualize basic research is important not only to define the concept but also to design policies that can effectively promote it. To this purpose, we asked a few questions to define the concept and goals of basic research. Q5 asked respondents to express their level of agreement with the following: “basic research can be defined as the research that is not intended to yield immediate practical benefits except for advancement of knowledge”. Survey participants had four options to choose: complete agreement, some agreement, some disagreement and complete disagreement. 32.5% of the respondents expressed complete agreement, 43.4% some agreement, 17.5% some disagreement and 6.6% complete disagreement (Figure 1A). To corroborate the responses of Q5, we designed a complementary question (Q6) to determine the level of agreement on the following: “basic scientists can ponder about the future indirect practical benefits of their research without losing their “basic status””. 71.2% of the respondents expressed complete agreement, 23.2% some agreement, 5.0% some disagreement and 0.7% complete disagreement (Figure 1B). The results of Q6 were therefore in agreement with the results of Q5.\n\n(A) Graph shows the levels of agreement to the following statement: “Basic research can be defined as the research that is not intended to yield immediate practical benefits except for advancement of knowledge”. 302 scientists answered the question; 2 skipped. (B) Graph shows the levels of agreement on the following statement: “Basic scientists can ponder about the future indirect practical benefits of their research without losing their “basic status””. 302 respondents answered the question; 2 skipped. (C) Graph shows how scientists answered the following question: “What should the most important goal of publicly funded basic BIOLOGICAL (not biomedical) research be?”. Surveyed scientists were given the indicated three choices. 297 answered the question; 7 skipped. (D) Graph shows how scientists answered the following question: “What should the most important goal of publicly funded basic BIOMEDICAL research be?” Surveyed scientists were given the indicated three choices. 301 answered question; 3 skipped.\n\nWe also designed two questions (Q9 and Q10) to understand what should be the “goals” of biological and biomedical research according to the basic scientists. In particular, Q9 asked respondents to answer the following question: “What should the most important goal of publicly funded basic BIOLOGICAL (not biomedical) research be?” Respondents had three options. 71.7% responded “pure advancement of knowledge, regardless of future applicability”, 21.9% responded “health benefit to the society (not necessarily in the near future)” and 6.4% responded that the most important goals are “other”, such as the “environmental or economical benefit to society” or “sustainability of our species and of the biosphere” (Figure 1C). Similarly, Q10 asked the following question: “What should the most important goal of publicly funded basic BIOMEDICAL research be?” Interestingly, we had a very different outcome. Only 8.6% of the respondents answered “pure advancement of knowledge regardless of future applicability”, while 85.7% answered “health benefit to society (not necessarily in the near future)” and 5.6% answered that the most important goals are “other” (Figure 1D).\n\nThus, these results clearly indicate that most scientists think that considering the indirect practical outcome of basic scientific investigations is compatible with the notion of basic research. In other words, basic research should not be conceptualized as being necessarily (or solely) driven by curiosity. Furthermore, these results suggest that scientists perceive the goals of “biological research” and “biomedical research” to be different, with a propensity to include health benefit to society as an important goal of biomedical research only.\n\nUnderstanding the motivations of people is important for designing policies that offer incentives to pursue certain goals. We therefore designed two questions to gather information on what motivates basic scientists. The respondents were asked to select their level of importance (“not a motivation”, “minimally important”, “moderately important”, “important” and “very important”) for six motivations. The rating average was then calculated after assigning a score from 1 to 5, to these five options. Scientists were therefore asked to provide feedback on the following (Q7) “the motivations of most basic biological/biomedical scientists are from:”. The rating average, for the motivation “pure advancement of knowledge, regardless of future applicability” was 3.91. The rating average for “health benefit to society (not necessarily in the near future)” was 3.93. The rating average for “gain of prestige” was 3.43. The rating average for “gain of money” was 2.42. The rating average for “satisfaction of their curiosity” was 4.24. The rating average for “satisfaction from solving puzzling problems” was 4.21 (Figure 2A).\n\n(A) Graph shows the responses of the surveyed scientists to the following input: “the motivation of MOST basic biological/biomedical scientists are from:”. Six different types of motivations were proposed. Respondents could rate each type of motivation as “not a motivation, “minimally important”, “moderately important”, “important” or “very important”. Rating averages for each type of motivation are also indicated (the scores were 1 to 5, from “not a motivation” to “very important”). 299 respondents answered the question; 5 skipped. (B) Graph shows the responses to the following input: “YOUR personal motivations as a scientist are from:”. Six different types of motivations were proposed. Respondents could rate each type of motivation as “not a motivation, “minimally important”, “moderately important”, “important” or “very important”. Rating averages for each type of motivation are also indicated (the scores were 1 to 5, from “not a motivation” to “very important”). 302 respondents answered the question; 2 skipped.\n\nTo see if scientists perceive themselves differently from other scientists, we also asked respondents to provide feedback on the following input (Q8): “YOUR personal motivations as a scientist are from:”. The rating average for “pure advancement of knowledge, regardless of future applicability” was 3.82. The rating average for “health benefit to society (not necessarily in the near future)” was 4.32. The rating average for “gain of prestige” was 2.79. The rating average for “gain of money” was 2.29. The rating average for “satisfaction of their curiosity” was 4.18. The rating average for “satisfaction from solving puzzling problems” was 4.16 (Figure 2B). Thus, these results show that, with the exception of “gain of money”, all other motivations are from “moderately important” to “very important” for more than 50% of the respondents. Moreover, these results show that scientists perceive themselves as more motivated by the pursuit of “health benefit to society (not necessarily in the near future)” and less motivated from the “gain of prestige” and “gain of money” than the average scientist.\n\nTo design policies to increase the practical impact of basic biomedical/biological research, it is first important to understand whether estimating the health benefit potential of basic research is in any way feasible, a topic that has being debated for many years9. We gathered feedback on this issue by asking respondents to express their level of agreement on scientists’ ability to estimate the potential future health benefits at different stages of the research process. Q11 stated: “Although it is difficult to assess the potential future health benefits to society from basic biological/biomedical research as described in written PROPOSALS, some degree of estimation is always possible”. 16.7% of the respondents were in complete agreement with this sentence, 57.7% in some agreement, 19.0% in some disagreement and 6.7% in complete disagreement (Figure 3A).\n\n(A) Graph shows the levels of agreement on the following statement: “Although it is difficult to assess the potential future health benefits to society from basic biological/biomedical research as described in written PROPOSALS, some degree of estimation is always possible”. 300 scientists answered the question; 4 skipped. (B) Graph shows the levels of agreement on the following statement: “Although it is difficult to assess the potential future health benefits to society from the RESULTS and FINDINGS of basic biological/biomedical research, some degree of estimation is always possible” 300 answered the question; 4 skipped.\n\nQ12 stated: “Although it is difficult to assess the potential future health benefits to society from the RESULTS and FINDINGS of basic biological/biomedical research, some degree of estimation is always possible”. 22.0% of the respondents were in complete agreement with this sentence, 61.0% in some agreement, 15.0% in some disagreement and 2.0% in complete disagreement (Figure 3B). These results therefore show that the majority (83%) of the surveyed scientists think that estimating the future health benefits to society from the proposals or outcome of basic biological/biomedical projects is realistically feasible.\n\nFunding agencies around the world commonly request that the potential health benefits of basic research projects are discussed in the written proposals. In order to understand what scientists think about this requirement, Q13 asked respondents to express their level of agreement in the following statement: “Written proposals about basic biological/biomedical research generally contain a section discussing potential future health benefits. These sections increase the likelihood that a project benefits future public health”. 12.3% of the respondents were in complete agreement with this statement, 35.0% were in partial agreement, 35.0% were in partial disagreement and 17.7% were in complete disagreement (Figure 4A).\n\n(A) Graph shows the levels of agreement on the following statement: “Written proposals about basic biological/biomedical research generally contain a section discussing potential future health benefits. These sections increase the likelihood that a project benefits future public health”. 300 scientists answered the question; 4 skipped. (B) Graph shows how surveyed scientists responded to the following question: “What percentage of public funding should be allocated to basic biological/biomedical research proposals in which discussing the potential of future health benefits to society is not required?”. 290 answered the question; 14 skipped. (C) Graph shows how surveyed scientists responded to the following question: “With regard to basic biological/biomedical research proposals in which discussing the potential of future health benefits to society is required, what average weight should be given to this potential in assigning scores for funding decisions?” 285 answered the question; 19 skipped.\n\nWe also proposed two questions with the purpose to shed light on how scientists would improve current funding criteria. In Q14, we asked to answer the following question: “What percentage of public funding should be allocated to basic biological/biomedical research proposals in which discussing the potential of future health benefits to society is not required?” According to the scientists of our sample, 41.6% of public funding, on average, should be allocated to research in which a discussion of the potential health benefits is not required in written proposals (Figure 4B) (standard deviation was 25.72; 3.4% of the respondents to this question declared 0%; 6.6% of the respondents declared 100%).\n\nIn Q15 we asked respondents to answer the following question: “With regard to basic biological/biomedical research proposals in which discussing the potential of future health benefits to society is required, what average weight should be given to this potential in assigning scores for funding decisions?” The average “weight” indicated by the scientists of our sample was 35.7% (Figure 4C) (standard deviation was 25.87; 6.7% of the respondents to this question declared 0%). Thus, this set of results indicates that the majority of scientists think that discussing the potential future health benefits in basic research proposals is not an effective way to increase the likelihood that a project benefits future public health. Interestingly, we noticed that principal investigators were significantly more in disagreement than post-docs (63.4% and 41.2%, respectively) with regard to the effectiveness of this policy in increasing societal benefits (Figure S5). Moreover, scientists believe that a considerable proportion of public funding (41.6%) should be allocated to research proposals in which discussing the future health benefits to society is not required.\n\nIn order to understand if scientists believe that motivational incentives could be more effective than stricter policies (such as the mandatory discussion of the potential medical benefits in research proposals), we asked (Q16) scientists to express the level of agreement on the following statement: “Motivational INCENTIVES, which are not based on restrictive policies such as the requirement to discuss the potential of future health benefits, CAN increase the degree to which basic biological/biomedical research is likely to benefit the future health of society”. With regard to financial incentives, 18.4% of the respondents were in complete agreement with this statement, 53.9% in some agreement, 16.0% in some disagreement and 11.6% in complete disagreement. With regard to non-financial incentives (e.g. awards, recognition), 13.5% of the respondents were in complete agreement with the statement, 60.6% in some agreement, 17.3% in some disagreement and 8.7% in complete disagreement (Figure 5A).\n\n(A) Graph shows the levels of agreement on the following statement: “Motivational INCENTIVES, which are not based on restrictive policies such as the requirement to discuss the potential of future health benefits, CAN increase the degree to which basic biological/biomedical research is likely to benefit the future health of society”. 293 answered the question; 11 skipped. The incentives were proposed either as financial or as non-financial. (B) Graph shows the levels of agreement on the following statement: “Motivational INCENTIVES, either “in addition to” or “in substitution of” restrictive policies, SHOULD be used to increase the degree to which basic biological/biomedical research is likely to benefit the future health of society”. 294 answered the question; 10 skipped. The incentives were proposed either as financial or as non-financial.\n\nTo understand if motivational incentives should be implemented and used, we also asked respondents (Q17) to express the level of agreement on the following slightly different statement: “Motivational INCENTIVES, either “in addition to” or “in substitution of” restrictive policies, SHOULD be used to increase the degree to which basic biological/biomedical research is likely to benefit the future health of society”. With regard to financial incentives, 15.4% of the respondents were in complete agreement with this statement, 47.1% in some agreement, 20.8% in some disagreement and 16.7% in complete disagreement. With regard to non-financial incentives, 14.1% of the respondents were in complete agreement with the statement, 55.7% in some agreement, 17.9% in some disagreement and 12.4% in complete disagreement (Figure 5B). Thus, these results suggest that the vast majority of basic scientists are in favor of motivational incentives (either financial or non-financial) to be used either “in addition to” or “in substitution of” more restrictive policies to increase the public health potential of basic biological/biomedical research.\n\nThe majority of the scientists who participated in the survey indicated that the most important goal of publicly funded basic biomedical research is the production of health benefits to the society (86%) (Figure 1D) and that the desire to effectively benefit society is an important or very important motivation for most of them (87%) (Figure 2B). While the benefits to society may be not realized in the near future, a substantial majority of respondents (74%) (Figure 3A) agreed/partially agreed on the idea that some degree of estimation of the potential contribution to human health is possible for every basic research proposal. Further, they indicated that, ideally, more than half of public funding should be allocated to proposals in which a discussion of the potential future health benefits to society is required. Moreover, with regard to the definition of basic research, nearly all respondents (94%) (Figure 1B) indicated that thinking about the future practical benefits of their research is compatible with the status of “basic” researchers, thus implying that basic research should not be conceptualized as (necessarily or solely) driven by curiosity.\n\nOur data also shed a light on scientists’ motivations (besides contributing to health benefit to society). This information is useful to design incentive-based policies. Our survey confirmed that the so-called “puzzle-motivation”—the satisfaction from solving puzzling problems—was an important motivator13 for almost all basic scientists (among our respondents 93% said that “satisfaction from solving puzzling problems” and 95% that “satisfaction of curiosity” were from “moderately” to “very important” motivations) (Figure 2B). The so-called “ribbon-motivation”—the gain of prestige and recognition—was significantly more important than the gain of personal money (among our respondents 60% said that the “gain of prestige” was a “moderately” to “very important” motivation for them compared to 41% who said the same for the “gain of money”) (Figure 2B). Moreover, the majority of respondents of the survey were in favor of using financial incentives (62%) and non-financial incentives (70%) to increase the degree to which basic biological/biomedical research is likely to benefit the future health of society (Figure 5B).\n\n\n\n\nDiscussion\n\nThe results of this survey provide valuable information to help conceive new effective policies to increase both the health benefit potential of basic biological and biomedical research and the work satisfaction of scientists without altering the nature and volume of scientific investigations (schematized in Figure 6). Building on these results, we conclude that nonfinancial soft incentives (nudges), in particular, are perceived as valuable tools to maximize the transformative value of basic research as they would not entail much work for scientists and can be implemented without significantly increasing public spending and bureaucratic burden. We also believe that soft incentives would be a valuable departure from current policies, which according to the scientists surveyed in our study, are ineffective. Indeed, despite 92% of respondents indicating that they are in favor of an increase in public funding for basic biological/biomedical research (Figure S4), a significant majority of the principal investigators (63%) (Figure S5) declared that the sections in written proposals aimed at discussing the potential future health benefits do not really increase the likelihood that a project will benefit future public health. Scientists also claimed that more public funding (on average the 42% of the total public funding committed to basic biological/biomedical research) should be devoted to basic biological/biomedical research proposals in which discussing the potential of future health benefits is not required (Figure 4B).\n\nIn order to increase the health benefits to society from basic “bio” research, policies should produce a good integration between basic and applied investigations and maximize transformative value of basic science. The directions of the arrows are intended only to express the capacity of transformative value of research (arrows pointing towards the top vertex of the triangle have maximum transformative value) and are not intended to reflect neither the quality of research nor the status of “basicness”.\n\nBased on these findings, we propose a few examples of policies based on soft (sometimes subconscious) incentives that could gently direct some scientists towards undertaking basic research inquiries with higher transformative value (Box 1). One example would be placing research laboratories in the proximity of hospitals to expose basic scientists to the view of patients and practicing physicians. Another one would be organizing more (non-mandatory) educational meetings in which scientists explain their work to the general public or to associations of patients (giving credit to the participating scientists). A similar proposal would be organizing periodical seminars inside research institutions to discuss the role of scientific research and scientists in society. Indeed we believe that effective policies would be the ones that exploit the scientist’s drive to achieve a good reputation and a role in benefiting society.\n\nLocating basic research laboratories in proximity of hospitals\n\nOrganizing educational meetings between scientists and the general public or patients associations. Acknowledging the participating scientists. Considering their participations during grant assignments, promotion, hiring etc.\n\nOrganizing more seminars (in academia and outside) about the role of scientific research and scientists in the society\n\nRecognizing the work of basic scientists in producing health benefits e.g. Requiring a list of seminal basic research articles for each new drug, medical device or other biological applications (see text for more details)\n\nPromoting more discussion on the concept and definition of basic research\n\nDifferent conceptualization of the notion of basic research (see text for more details)\n\nA model of such an incentive would be to formally recognize the basic scientists when new drugs or medical devices are approved, as we recently proposed14. This type of incentive would make use of the “ribbon-motivation” but without undermining the “puzzle-motivation” or research freedom generally13. This system could work by implementing a “bibliography of basic papers” for each newly approved drug. To apply this idea, a peer review group would identify the basic papers that have been influential for the development of the drug (or other biological applications) or, alternatively, review a list proposed by the drug owner14. A list of fifty to one hundred basic research papers would be selected and appear in the public databases (such as the Orange Book of the FDA) and in the drug package. This system would be a “weak attractor” because it would not distract scientists from basic research but it would represent a small, mostly unconscious, incentive to pursue research lines that can more easily lead to future drugs. Therefore this system would not dramatically affect the whole “ecosystem” of the scientific research that indeed needs to be made of a balanced mix of the different types of research, from the “purely” basic to the “purely” applied (Figure 6). This method would also present the advantage of increasing public awareness of the role of basic science, which we think is often underestimated by lay people as well as politicians.\n\nMoreover, we believe that a different conceptualization of the notion of basic research would help in increasing the transformative values of fundamental investigations. A portion of basic research should (continue to) be devoted to purely curiosity-driven purposes as knowledge per se has a value and increases the quality of life of people through fascination and ‘soul nourishment’. However, basic research should not be conceptualized as solely driven by curiosity. Indeed, in our survey, nearly all respondents (94%) (Figure 1B) indicated that thinking about the future practical benefits of their research is compatible with the status of “basic” researchers. Therefore, similar to the Organisation for Economic Co-operation and Development (OECD)’s division of the continuum of basic research into pure basic research and oriented basic research (http://stats.oecd.org/glossary/detail.asp?ID=192), we believe basic research can usefully be divided into two broad categories: solely curiosity-driven research and research driven by a vision of how the knowledge generated might be useful for future applications. In this context, the term “blue skies research”, sometimes used to define the entire field of basic research15 might be used for those studies that are solely (or largely) curiosity-driven. Even if curiosity does remain one of the main motivators for conducting and studying science, we believe that basic research should be conceptualized as research that focuses on basic mechanisms of natural phenomena rather than research that is intended to satisfy scientists curiosity (as it is frequently presented in the mass media). Along these lines, we must also revisit the idea that since the future benefits of basic research cannot be accurately predicted, all basic research is equally valuable, i.e. every imaginable basic investigation would have the same exact potential of practical outcome. In fact, the great majority of scientists who took part in our survey pointed out that, despite the fact that it is usually necessary to undertake a very long pathway (the “countless sources” mentioned by Abraham Flexner3) before being able to funnel basic knowledge toward more applicative studies, some degree of assessment of the transformative value of basic investigations is always possible. It follows that since the potential benefits for society are roughly predictable, basic research can be evaluated prospectively; this does not lessen the “basic status” of either the research or the scientist. Such a revised mindset could “nudge” more basic scientists (and grant funders) to wonder about the future impact of their investigations.\n\nBasic research advances knowledge that, regardless of its possible practical outcome, has a value per se. In addition, basic research has also the potential to produce more practical benefits to humanity, such as the prevention and treatment of diseases. As a society, we have the moral obligation to try to maximize this potential. We believe, and the data presented in the paper support, the idea that soft incentives can be valuable tools for increasing this potential without corrupting the spirit of fundamental investigations, thus further aligning the goals of cell and molecular biologists with those of the broader public health community.\n\n\nMethods\n\nEthics statement: On April 2, 2012, the Institutional Review Board (IRB) of Harvard School of Public Health determined that the proposed study meets the criteria for exemption per the regulations found at 45 CFR 46.101(b)2. The IRB made the following determinations: Research Information Security Level; the research is classified, using Harvard's Data Security Policy, as Level 1 data. The notification was signed by QA/QI specialist.\n\nThe survey was designed as an online questionnaire (powered by SurveyMonkey, www.surveymonkey.com) made of 17 questions (Q1–Q17) plus one additional field for free comments. Answers could be provided through multiple choices or, alternatively, textboxes for alphanumerical entries. Each single question had the option to be skipped. The survey was sent to a sample of scientists involved in basic biological/biomedical studies (for the most part, cell and molecular biology studies). The scientists were also asked to confirm their level of involvement in basic fundamental research (see results section). The responses were collected during 9 consecutive weeks during 2012 (end of April to the end of June). Principal investigator (PI) scientists were contacted by email after consulting the websites of Harvard University and some affiliated institutes (Brigham and Women’s Hospital, Beth Israel Deaconess Medical Center, Dana-Farber Cancer Institute, Joslin Diabetes Center and Children’s Hospital); the majority of principal investigators were asked to forward the survey to members of their own groups. Post-docs were contacted either by their PIs or by using university-associated mailing lists and networking. Also a few scientists with other types of position (e.g. PhD students, instructors, research assistants) took part in the survey, generally contacted by their PIs. In addition to the specific request to forward the survey to their own groups or to close intra-institutional colleagues, the contacted scientists were specifically asked not to forward the survey to the outside community. The survey was completely voluntary and anonymous.\n\n\nData availability\n\nFigshare: Responses of Harvard Medical School (and affiliate) scientists to an online survey on basic and biomedical science policy, http://dx.doi.org/10.6084/m9.figshare.90283716.", "appendix": "Author contributions\n\n\n\nAndrea Ballabeni, Andrea Boggio and David Hemenway designed the study and wrote the manuscript. Andrea Ballabeni collected the data.\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\nWe are grateful to Daniel Wikler (Harvard School of Public Health), Alessandro Doria (Joslin Diabetes Center), Marc W. Kirschner (Harvard Medical School), Chiara Manzini (Children’s Hospital) and Paul H. Lerou (Brigham and Women’s Hospital) for their valuable feedback during the preparation of the survey. We thank Daniel Sarewitz (Arizona State University) and Thomas Stossel (Brigham and Women’s Hospital) for helpful feedback during manuscript preparation.\n\n\nSupplementary materials\n\n\n\n\nReferences\n\nSampat BN: Mission-oriented biomedical research at the NIH. Research Policy. 2012; 41(10): 1729–1741. Publisher Full Text\n\nLim SS, Vos T, Flaxman AD, et al.: A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380(9859): 2224–2260. PubMed Abstract \| Publisher Full Text\n\nFlexner A: [The usefulness of useless knowledge]. Med Deporte Trab. 1952; 17(117): 5274–5278. PubMed Abstract\n\nFreeland RM: Academia's Golden Age: Universities in Massachusetts, 1945–1970. (Oxford University Press, New York). Reference Source\n\nMcClure SM, Laibson DI, Loewenstein G, et al.: Separate neural systems value immediate and delayed monetary rewards. Science. 2004; 306(5695): 503–507. PubMed Abstract \| Publisher Full Text\n\nHemenway D: Why we don't spend enough on public health. N Engl J Med. 2010; 362(18): 1657–1658. PubMed Abstract \| Publisher Full Text\n\nBisias D, Lo AW, Watkins JF: Estimating the NIH efficient frontier. PLoS One. 2012; 7(5): e34569. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCollins FS, Green ED, Guttmacher AE, et al.: A vision for the future of genomics research. Nature. 2003; 422(6934): 835–847. PubMed Abstract \| Publisher Full Text\n\nBornmann L: Measuring the societal impact of research: research is less and less assessed on scientific impact alone--we should aim to quantify the increasingly important contributions of science to society. EMBO Rep. 2012; 13(8): 673–676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrewer GJ: Fundamental problems lie ahead in the drug discovery and commercialization process: restructuring of the pharmaceutical industry and an improved partnership with academia are required. J Investig Med. 2006; 54(6): 291–302. PubMed Abstract \| Publisher Full Text\n\nThursby J, Thursby MC: Has the Bayh-Dole act compromised basic research? Research Policy. 2011; 40(8): 1077–1083. Publisher Full Text\n\nThaler RH, Sunstein CR: Nudge: improving decisions about health, wealth, and happiness. New Haven: Yale University Press.2008. Reference Source\n\nLam A: What motivates academic scientists to engage in research commercialization: ‘Gold’, ‘ribbon’ or ‘puzzle’? Research Policy. 2011; 40(10): 1354–1368. Publisher Full Text\n\nBallabeni A, Boggio A, Hemenway D: Recognizing Basic Science Contributions. Scientist. 2014; 28(1): 26–27. Reference Source\n\nLinden B: Basic Blue Skies Research in the UK: Are we losing out? J Biomed Discov Collab. 2008; 3: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBallabenia A, Boggio A, Hemenway D: Responses of Harvard Medical School (and affiliate) scientists to an online survey on basic and biomedical science policy. Figshare. 2014. Data Source" }	[ { "id": "3242", "date": "30 Jan 2014", "name": "Giorgio Scita", "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 well-conducted study that clarifies the way basic science is perceived by basic scientists.It would have been interesting to assess whether the career stage of the respondents has any impact on their answers to the survey.In the discussion a point is raised as to whether there is the need to revisit “the idea that since the future benefits of basic research cannot be accurately predicted, all basic research is equally valuable” based on the finding that the majority of scientists believe that “some degree of assessment of the transformative value of basic investigations is always possible”. However in the absence of a clear definition of what the degree of assessment is, it remains difficult to fully support such a perception. Thus caution should be exerted in drawing a conclusion as to whether this mindset could “nudge” more basic scientists (and grant funders) to wonder about the future impact of their investigations.The list of soft motivational incentives in BOX1 is thought provoking and stimulating.", "responses": [ { "c_id": "847", "date": "03 Jun 2014", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would like to thank Dr Giorgio Scita for his helpful feedback and for giving us the possibility to improve our manuscript. We have tried to address each of his comments. We provide below our answers point-by-point. - It would have been interesting to assess whether the career stage of the respondents has any impact on their answers to the survey.To address this comment we have created a new paragraph in the Results section in which we list the most meaningful differences between the answers of the principal investigators and the answers of the post-docs. In general, there were not huge differences. However we believe that a few differences are noteworthy.We have also added a few sentences in the Discussion section in order to comment about these differences e.g. “ … The use of “nudges” seems to be particularly promising with the basic scientists at the earlier stages of their career; compared to principal investigators, post-docs are more likely to think that the major goal of basic biological and biomedical research is to provide health benefits to society, are even more driven by prestige and financial motivations and are even more in favor of the use of soft incentives…”.We have also submitted the summaries of all the answers of the principal investigators and all the answers of the post-docs; we propose to insert two links into the main text to enable access to these two summaries. - In the discussion a point is raised as to whether there is the need to revisit “the idea that since the future benefits of basic research cannot be accurately predicted, all basic research is equally valuable” based on the finding that the majority of scientists believe that “some degree of assessment of the transformative value of basic investigations is always possible”. However in the absence of a clear definition of what the degree of assessment is, it remains difficult to fully support such a perception. Thus caution should be exerted in drawing a conclusion as to whether this mindset could “nudge” more basic scientists (and grant funders) to wonder about the future impact of their investigations. The referee is right about the absence of a clear definition of what a “degree of assessment” is. For this reason we have added some thoughts in the Discussion section: “Of course, it will also be important to estimate in the best possible way and case by case, the degree (i.e. possibility) of assessment of the transformative value as an overestimation of our ability to assess the transformative value of research projects could have negative effects on both the innovation potential and the scientist satisfaction. Plausibly, future study and discussion will shed more light on these concepts and increase our ability to assess the social potential of fundamental investigations.”We hope that this addition will fully address the referee comment." } ] }, { "id": "4134", "date": "31 Mar 2014", "name": "Daniel Strech", "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 addresses an interesting and relevant topic. Box 1 in particular is innovative and makes a good link to the findings. I have the following comments:More remarks are needed about the relevant body of evidence. Are there other studies that investigated the self-understanding of basic scientists in biomedicine? The Background mentions none and the Discussion only one paper. The main text should make clear that the response rate and reasons for non-response are not known. A brief explanation together with a justification on why this approach was selected should be given. Figure 6 is somewhat simplistic and packed with several unjustified assumptions on what the (true) positive and negative effects of specific pattern of resource allocation to basic or translational research are. This should be deleted. The legend to figure 6 is very general, for example the recommendation saying that one should “produce a good integration between basic and applied investigations …”.", "responses": [ { "c_id": "848", "date": "03 Jun 2014", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would like to thank Dr Daniel Strech for his helpful feedback and for giving us the possibility to improve our manuscript. We have tried to address each of his comments. We provide below our answers point-by-point. - More remarks are needed about the relevant body of evidence. Are there other studies that investigated the self-understanding of basic scientists in biomedicine? We are really grateful for the excellent advice of providing more references (and connected remarks) about the relevant body of knowledge. We have therefore added 28 new references and related new remarks in the Introduction and in the Discussion sections. - The main text should make clear that the response rate and reasons for non-response are not known. A brief explanation together with a justification on why this approach was selected should be given. We agree with the referee that this is an important thing to discuss in the manuscript. Therefore we have added a discussion in the Methods section:“The response rate is not known because we do not know how many scientists actually read the invitation email and how many principal investigators forwarded the invitation to their lab members. We used this approach because we wanted to maximize the sample size. By taking in consideration only the scientists that took part to the survey, the response rate was very high for all the questions (on average, fewer than 2% of the respondents, with a range from 0% (Q2 and Q3) to 6.3% (Q15) skipped any of the 17 questions); this suggests that the survey did not contain difficult-to-understand or difficult-to-answer questions. Therefore the decision of participating (or not) to the survey was probably not based on the nature of the questions but rather on other factors (e.g. lack of time) that, conceivably, have only a marginal effect on the representativity of the sample. Moreover, the survey was completely anonymous. For these reasons, we believe that the sample of scientists that took part to this survey is fairly representative of the entire population of scientists working in the same setting (Harvard Medical School and affiliated institutes).” - Figure 6 is somewhat simplistic and packed with several unjustified assumptions on what the (true) positive and negative effects of specific pattern of resource allocation to basic or translational research are. This should be deleted. The legend to figure 6 is very general, for example the recommendation saying that one should “produce a good integration between basic and applied investigations …”. We thank the referee for this comment. While figure 6 (now figure 8) may look simplistic and general, this design is deliberate because we think that the simple schematizations contained in it help the readers visualize and fully grasp some of the concepts discussed in the manuscript. For these reasons, we would like to keep this figure. However, we have slightly edited the figure legend and the figure itself in order to further emphasize the concept that we are just showing a schematic ‘working model’ (and like all working models, it may contain assumptions and generalizations)." } ] } ]	1	https://f1000research.com/articles/3-20
https://f1000research.com/articles/3-7/v1	13 Jan 14	{ "type": "Research Article", "title": "Epiphyte response to drought and experimental warming", "authors": [ "Joshua M. Rapp", "Miles R. Silman", "Miles R. Silman" ], "abstract": "The high diversity and abundance of vascular epiphytes in tropical montane cloud forest is associated with frequent cloud immersion, which is thought to protect plants from drought stress. Increasing temperature and rising cloud bases associated with climate change may increase epiphyte drought stress, leading to species and biomass loss. We tested the hypothesis that warmer and drier conditions associated with a lifting cloud base will lead to increased mortality and/or decreased recruitment of epiphyte ramets, altering species composition in epiphyte mats. By using a reciprocal transplant design, where epiphyte mats were transplanted across an altitudinal gradient of increasing cloud immersion, we differentiated between the effects of warmer and drier conditions from the more general prediction of niche theory that transplanting epiphytes in any direction away from their home elevation should result in reduced performance. Ramet mortality increased, recruitment decreased, and population size declined for epiphytes in mats transplanted down slope from the highest elevation, into warmer and drier conditions, but epiphytes from lower elevations showed greater resistance to drought in all treatments. Epiphyte community composition changed with elevation, but over the timescale of the experiment there were no consistent changes in species composition. Our results suggest some epiphytes may show resistance to climate change depending on the environmental context, although if climate change results in consistently drier conditions and higher cloud bases, biomass loss and shifting species composition in epiphyte communities is likely.", "keywords": [ "Andes", "cloud immersion", "reciprocal transplant", "cloud base", "altitudinal gradient" ], "content": "Introduction\n\nTropical montane forests, often referred to as cloud forests, harbor high species diversity, provide water and protect water quality for numerous people in tropical countries, and are under particular threat from climate change1–3. Most cloud forest regions of the world, including the tropical Andes, are considered hotspots of biological diversity4, and plant species endemism often reaches high levels within cloud forest5,6. Epiphytes, non-parasitic plants that depend on other plants for support and are not in contact with terrestrial soil, are key components of cloud forest biodiversity and play critical roles in the hydrological and nutrient cycling of montane ecosystems. Not only can vascular epiphytes make up 30 percent or more of plant species diversity in tropical montane forests7, they also provide keystone resources for birds, insects, and other animals8–11. Through cloud stripping, epiphytes increase total moisture captured by forest canopies12–14, and are important in nutrient cycling15,16. Cloud immersion is important for many epiphyte species to maintain a positive water balance and avoid desiccation17. Because of their sensitivity to moisture levels, epiphytes are considered indicator species in cloud forests for changing water balance conditions18, particularly those in the wet tropics19.\n\nOn a typical tropical mountain, where temperature decreases with altitude, there is a gradient of increasing cloud incidence with altitude. Cloud formation is dependent on the vapor content of air and air temperature, both of which are predicted to change with global warming1. Atmospheric moisture levels are much less easily predicted than temperature in climate models, but a multi-model ensemble of climate simulations showed a trend towards drying in many tropical regions20. Climate model projections in the Andes include warmer temperatures and lower precipitation in the dry season21–23, which could place cloud forest plants under increased drought stress. Cloud base height is also predicted to rise, due to a combination of higher temperature and lower atmospheric moisture input by vegetation due to lowland deforestation and reduced transpiration because of increased atmospheric CO224–26. In Costa Rica, an increase in the elevation of cloud base has been demonstrated and has already lead to the extinction of cloud forest species27, although this may have been associated with a severe El Niño event in 1986–87 rather than a long term drying trend28.\n\nThe sensitivity of vascular epiphytes to changes in cloud incidence was demonstrated experimentally in Monteverde, Costa Rica, where vascular epiphytes transplanted below the cloud base had shorter lifespans and higher leaf mortality29. It is not clear, however, whether this result can be generalized to continental cloud forests such as the eastern Andes. Cloud forests vary worldwide, with differences in cloud base height and the proportion of moisture received by the vegetation via cloud stripping versus rainfall30,31. Cloud forests near coastlines are heavily influenced by ocean conditions32, while cloud formation on continental mountain ranges is dependent on moisture flux across continents driven by synoptic weather patterns. Deforested areas in Costa Rica have fewer clouds than adjacent forested areas33, but simulations suggest that sea-surface temperature has a greater impact on lifting condensation level than deforestation34. Cloud forests in continental mountains like the Andes are expected to be more sensitive to conditions of the adjoining lowland ecosystems, particularly deforestation25,32. In addition, complex topography in the eastern Andes and its interaction with prevailing winds leads to wet and dry areas within regions broadly considered cloud forest35. Given the diversity of cloudiness and precipitation regimes that epiphytes as a group are exposed to, it is reasonable to expect that epiphytes may be adapted to local moisture regimes. For instance, epiphytes in lowland dry or seasonal forest have high desiccation tolerance36,37. Epiphytes in continental cloud forests like the Andes, which experience variable cloudiness regimes, may have more resistance to drought than those in locations with more stable cloud bases. Likewise, epiphytes growing at lower elevations, below the cloud base, may have greater drought tolerance than epiphytes growing above the cloud base.\n\nBeginning in the 2005 austral winter (June and July), we conducted a year-long reciprocal transplant experiment across an elevational gradient in cloud formation in the eastern Andes of southern Peru to test the effect that cloud immersion has on the performance of vascular epiphytes. The reciprocal transplant design also allowed us to distinguish between the effects of moving mats away from their home elevation versus moving plants into lower moisture conditions. This is a key control, not often made climate change transplant studies [e.g.29,38], for if epiphytes are locally adapted, reduced performance is expected if moved in any direction from their bioclimatic optimum. The specific questions we addressed were: (1) Does demographic performance decline when epiphytes are moved farther from their home elevation? (2) Is this effect greater when transplanted down-slope into drier and warmer conditions as predicted by results from Costa Rica?29 (3) Do epiphyte species of different functional types (e.g. strap-leaf ferns, orchids, ericaceous shrubs) respond similarly to moisture gradients?\n\nWe focused on three aspects of demographic performance, ramet survival, recruitment, and change in population size. Ramet survival allows us to examine treatment effects on existing individuals, while ramet recruitment and population change gives insight into treatment effects on epiphyte populations. We expected that ramet survival and recruitment would decrease and population change become negative as mats were moved farther from their home elevation. We also hypothesized that the decrease in ramet survival and recruitment would be greater for mats moved down-slope than for mats moved upslope if moisture level is the dominant factor in determining epiphyte species distributions. A priori, we expected different functional types to respond similarly to the altitudinal gradient such that any transplant effect would be similar across groups. Specifically, we expected functional types to have differing ramet mortality, but that these would be balanced by ramet production such that species composition would remain stable in undisturbed mats.\n\n\nMaterials and methods\n\nThe Kosñipata Valley (13°03′S, 71°33′W) lies along the eastern slope of the Andes in southern Peru. Elevations range from about 800 m to over 4000 m, and vegetation changes from pre-montane rainforest at the lowest elevations to tropical subalpine forest and puna (alpine grassland) at the highest elevations39. The experiment was installed along a single forested ridge, with three transplant sites at 1500, 1650, and 1800 m. We chose these elevations because the large increase in vascular epiphyte and bryophyte biomass40, and step changes in soil properties41 and biomass carbon stocks42 between 1500 and 2000 m elevation in the Kosñipata Valley are likely associated with higher cloud incidence and lower temperatures, as seen on other tropical mountains43,44. The bedrock underlying the ridge is Permian granite, and soils are classified as umbric Gleysols45.\n\nThe Kosñipata Valley has a perhumid climate described in detail in Rapp and Silman46. Temperature decreases linearly with altitude and annual rainfall is high across the gradient, with a distinct, but weak dry season (monthly rainfall > monthly potential evapotranspiration except during drought). Vapor pressure deficit (VPD) typically decreases with altitude across the east Andean slope above 1500 m for all months except June46. While differences in VPD between 1500 m (lowest experimental elevation) and 1800 m (highest experimental elevation) are typically small, excursions to higher VPD (greater desiccation) are more severe at 1500 m46. The dry season (May – August) is when cloud immersion is expected to be most important for cloud forest plants. In July, the driest month, high relative humidity associated with cloud immersion (> 95%) is more common at 1840 m than at 1500 m at climate stations approximately 1 km from the study site, while vapor pressure deficits greater than 1.0 kPa are more common at the higher elevation (Table 1). Vapor pressure deficits greater than 1.0 kPa are associated with moisture stress in cloud forest plants47,48. Daily photographs during installation of the experiment in June and July 2005 and July 2006, confirmed that cloud base height did not differ between years (Χ2 = 3.32, p = 0.19), and that cloud frequency increased with elevation (cloud base < 1500 m in 9% of observations; 1500 – 1800 m in 30% of observations; > 1800 in 61% of observations).\n\nValues represent the mean for July in 2007, 2008, and 2009, except for precipitation, which does not include data for 2007. RH95 is the proportion of time relative humidity was greater than 95%. VPD excursions is the number of days per month in which vapor pressure deficit (VPD) was 1.0 kPa or greater.\n\nIn the 2005 austral winter (dry season: June and July), we selected five Alzatea verticillata Ruiz & Pav. trees at each of three elevations, 1500 meters, 1650 meters, and 1800 meters elevation. Alzatea was an appropriate choice for a host tree because: (1) it was common at all three elevations; (2) it attains large size and has strong wood suitable for supporting climbers working in the trees; and (3) its unique architecture resulted in many large horizontal branches that supported sizeable epiphyte mats. We accessed trees using roped arborist techniques49. In each tree, we chose four sections of epiphyte mat that were at least 25 cm wide and 30–40 cm long. Within each of the 60 mats, we marked an area of 25 × 25 cm with wire, and marked all ramets of vascular epiphytes within it, identified them to morpho-species, and recorded the length of shoots and number of leaves of each ramet. Nonvascular epiphytes were present, but not considered in this experiment. Most mat dwelling epiphytes are clonal, with individual ramets connected by subsurface stems, but capable of surviving without connection to other ramets. It was difficult to determine individual genets without excavating the plants, so we identified and measured individual ramets rather than genetically distinct plants. For strap-leafed ferns in the genus Elaphoglossum, each ramet was a single leaf, while ramets for other species consisted of one or more stems with multiple leaves. On each tree, one of the mats was left in place to serve as an undisturbed control. We cut each of the other three mats from the tree, lowered them to the ground, and then transplanted one each to a random tree at each of the three elevations. After all transplants were complete, each tree had one undisturbed mat, one mat that had been removed and then replaced at the same elevation, and one mat from each of the other two elevations. We tied each mat in place using wire, and then watered it with one liter of water to minimize any desiccation effect that handling may have had. Supplementary Figure 1–Supplementary Figure 4 illustrate the process of transplanting the epiphyte mats.\n\nWe left the mats undisturbed for one year, and resurveyed them the following year in June and July 2006 (Dataset). We searched for all marked ramets and counted the new ramets of each morpho-species. We assumed ramets obviously more than a year old (28 out of 1400+ original ramets) had lost their tag if previously marked ramets of the same species were not found in the same mat. Eight ‘old’ ramets were still not accounted for; we assumed these were missed during the first census. Any other missing ramets were assumed to be dead.\n\nWith these data, we defined three measures of population performance: 1) survival, 2) recruitment, and 3) population change. Survival was defined as:\n\nSurvival = (N2005 – D2006) / N2005,\n\nwhere N2005 was the number of ramets surveyed in a mat in 2005, and D2006 was the number of ramets surveyed in 2005 that had died by 2006. Recruitment was defined as:\n\nRecruitment = n2006 / N2005,\n\nwhere n2006 was the number of new ramets surveyed in 2006, which were not present in 2005. Population change was defined as:\n\nPopulation change = (N2006 – N2005) / N2005.\n\nAll analyses were performed using the mat as the experimental unit to account for within-mat correlations between ramets, i.e. to avoid pseudo replication. We fitted models to data that included all ramets irrespective of species to explore the overall community patterns of ramet survival, recruitment, and population change, and then modeled common species separately to look at individual species responses. The common species occurred in at least half (10) of all epiphyte mats transplanted from at least one elevation. These included four common morpho-species identified to genus, by which we will refer to them: a strap-leaf fern (Elaphoglossum Schott ex J. Sm.), two orchid species (Maxillaria Ruiz & Pav.; Scaphyglottis Poepp. & Endl), and an ericaceous shrub (Cavendishia Lindl.). Collectively, these species accounted for 78% (1127/1452) of the ramets surveyed in the initial 2005 survey (Table 2).\n\nNumbers in parentheses indicate number of mats the ramets were found in. Bold indicates ramets used in single species analyses.\n\nWe tested whether ramet recruitment was different than mortality in transplanted mats using a two-sided t-test. We then analyzed ramet survival, recruitment, and population change with respect to experimental manipulations using generalized linear mixed-effects models (GLMMs). Survival was modeled as a binomial distribution with a logit link function to account for the binary nature of the response (alive, dead). Recruitment (new ramets in 2006) and population change (total ramets in 2006) were modeled as a rate relative to the initial ramets per mat by using a Poisson distribution with a log link, and adding an offset of the log of the number of initial ramets in 2005. To account for the natural blocking by tree in our experimental design, models that included multiple source elevations and transplant elevations also included random effects for source tree and transplant tree. Models including only one source elevation included a random effect for source tree only. Likelihood ratio tests were used to assess the fixed effects, while Wald z-tests were used to evaluate differences between levels of fixed effects. We did not evaluate the significance of random effects because they were a required part of our experimental design. Finally, we confirmed that the residuals of the final model were not overdispersed50 using code from Bolker et al.51. All analyses were done in R [Version 2.15.2;52]. In all analyses we considered an effect significant if the P-value was less than 0.05.\n\nFirst, we tested for an effect of manipulating mats using data for undisturbed control mats and mats transplanted within elevation. Source elevation and treatment (transplant versus undisturbed) were modeled as fixed effects in this analysis. Then, we tested for effects of source and transplant elevation on the response variable, using data from just the transplanted mats. We took this two-tiered approach because a full model including all mats was unbalanced (e.g. there could not be a control mat that moved between elevations) and statistical models accounting for this would not converge computationally.\n\nTo investigate patterns in mat species composition we used Detrended Correspondance Analysis (DCA) because our compositional data collected across a directional gradient matched the assumptions of DCA. First, we investigated the change in composition versus elevation using the pre-transplantation composition of all mats. We then investigated compositional change due to experimental treatments by ordinating the composition of all mats during 2005 before transplantation, with the composition of mats in 2006, one year after transplantation. Permutation Multivariate Analysis of Variance using distance matrices [function adonis in the vegan R package;53] was used to test for compositional changes with altitude and among years due to the transplantation.\n\n\nResults\n\nFirst, we tested for an effect of transplantation independent of elevational distance moved by asking whether epiphytes in mats transplanted to the same elevation had different ramet survival, recruitment, and population change or turn-over than those in intact mats. Across all species, there was no significant effect of transplant, elevation, or their interaction on survival, recruitment, or population change of epiphyte ramets (Table 3, Figure 1). However, individual species were affected by transplantation. For Elaphoglossum, survival was lower in mats transplanted to another site at the same elevation than in undisturbed controls, but there was no effect of elevation on survival or any interaction between elevation and transplantation (Table 4, Figure 1). There was an interaction between elevation and transplantation for recruitment and population change in Elaphoglossum, however (Table 4); both were lower for transplanted mats at 1500 m and 1650 m, but higher for mats transplanted at 1800 m (Figure 1). For Maxillaria, recruitment was lower in transplanted mats, but not affected by elevation, and neither survival nor population change was affected by either transplanting or elevation (Table 4, Figure 1). For Cavendishia, recruitment and population change were lower for transplanted mats (Figure 1), but only significantly so for recruitment; survival was unaffected by transplantation (Table 4). Transplanting did not affect survival, recruitment, or population change in Schaphyglottis (Table 4, Figure 1).\n\nPoints show survival (top), recruitment (middle), and population change (bottom) of individual mats in 2006, each expressed as a proportion relative to the number of ramets present in 2005. Thick horizontal lines and boxes depict the modeled mean and 95% confidence intervals, respectively. White shading depicts controls, dark shading transplant.\n\nSE: standard error.\n\n[R] indicates random effect.\n\nSE: standard error.\n\n[R] indicates random effect.\n\nAcross all species, there were no significant effects on survival of any of the treatments for mats transplanted across elevations (Table 5 and Figure 2). For recruitment and population change, there was a significant interaction between source and transplant elevation (Table 5), with both positively associated with elevation for mats transplanted from 1500 and 1800 m, but negatively associated with altitude for mats from 1650 m (Figure 2). Overall for transplanted mats, more ramets died than were recruited (mean change number of ramets per mat between years = -1.38; two-sided t-test, P = 0.01).\n\nPoints show survival (top), recruitment (middle), and population change (bottom) of individual mats in 2006, each expressed as a proportion relative to the number of ramets present in 2005. Thick horizontal lines and boxes depict the modeled mean and 95% confidence intervals, respectively. White shading depicts 1500 m elevation, light grey 1650 m elevation and dark grey 1800 m elevation.\n\nSE: standard error.\n\n[R] indicates random effect.\n\nElaphoglossum ramets in mats originating at 1500 m had consistently and significantly higher survival than those originating at 1650 m or 1800 m, but there was no effect of transplant elevation on survival (Table 6, Figure 2). For both recruitment and population change, however, there was a significant interaction between source elevation and transplant elevation, with both recruitment and population change declining in mats transplanted at lower elevations for mats originating at 1500 m and 1800 m, but for mats originating at 1650 m recruitment was greater and population change more positive for mats transplanted to 1500 m than for mats transplanted to higher elevation (Table 6, Figure 2).\n\nSE: standard error.\n\n[R] indicates random effect.\n\nFor Maxillaria, the only significant effect for transplanted mats was that for transplant elevation on recruitment (Table 6); recruitment was low in all transplanted mats, but there was zero recruitment in mats transplanted to 1500 m (Figure 2). There were no significant effects of source elevation or transplant elevation on survival or population change (Table 6), but survival was lower and population change more negative for ramets transplanted to 1500 m (Figure 2).\n\nAll three measures of performance were unaffected by transplant elevation in Cavendishia (Table 6, Figure 2). For Schaphyglottis, survival, recruitment, and population change were all progressively lower in mats transplanted to lower elevations (Figure 2), but the difference was significant only for survival (Table 6).\n\nPrior to transplanting mats, the epiphyte community composition showed significant differences across the elevational gradient, although relatively little of the variation could be explained by elevation (Table 7); most of the compositional separation was between mats at 1500 m and the other two elevations (Figure 3). Morpho-species richness increased with elevation (Poisson regression, Z = 2.446, P = 0.0144), while the number of ramets per mat declined (Poisson regression, Z = -2.281, P = 0.0225; Table 8). Comparison of pre- and post-treatment species compositions in mats revealed no directional shift in community composition due to transplantation (Table 7). A few mats did show large changes (Figure 3), likely because of large changes in abundance in Elaphoglossum, either through high ramet mortality or recruitment (Figure 2).\n\nArrows connect the compositions of individual mats before and a year after transplantation. Line width depicts direction of transplanting. Hulls are drawn around the 2005 composition of mats that originated at the same elevation, and labels are placed at the hull centroid.\n\nSignificance levels: **p < 0.01.\n\nPer mat values are means with standard error in parentheses.\n\n\nDiscussion\n\nVascular epiphytes transplanted down slope from our highest elevation had lower demographic performance when transplanted to the lowest elevation, suggesting warmer temperatures and lower cloud immersion will cause community-level changes for species currently above the cloud base. This result corroborates previous work in another tropical montane site, which found fewer leaves and shorter life-spans for vascular epiphytes moved down slope29. However, reciprocal transplants between all elevations revealed unexpected dynamics, with demographic rates differing in their response and species responding individualistically to the treatments, perhaps reflecting functional type differences (Figure 2). Cavendishia, a small woody shrub showed the least response, while Elaphoglossum, a strap-leafed fern in which individual leaves were the measurement unit was most responsive to treatments. The two orchid species were intermediate, with Schaphyglottis responding more strongly than Maxillaria which has stouter stems. Thus, functional differences in ramet construction may account for the differences in response among species to the elevational transplants. In general, it appears that epiphytes responded to water stress but we also found evidence for local adaptation. This response to transplanting was strongest in those transplanted from the highest elevation, which has the highest degree of cloud immersion. Epiphytes from lower elevations only benefitted slightly from increased water availability, possibly indicating they are better adapted to withstand drought stress. Epiphytes from the middle elevations, where cloud immersion is intermediate, responded in more idiosyncratic ways to transplantation. Finally, while composition changed across the elevational gradient, there was no significant directional shift in composition due to any of the transplant treatments (Figure 3, Table 7). The relative resistance of epiphytes to expected transplant-induced moisture stress found in this study could be due to two competing factors, described below in more detail: (1) rainfall compensation in this pluvial system, where high rainfall is sufficient to maintain epiphyte water-balance below cloud base; and/or (2) higher epiphyte drought tolerance from a history of variable rainfall and occasional drought in these continental mountains.\n\nWhile our results for epiphytes transplanted from the highest elevation are consistent with the hypothesized altitudinal gradient in moisture stress, this gradient had less of an effect on epiphyte performance than the one in Monteverde, Costa Rica29. The relative importance of cloud immersion for the distribution of epiphytes in this system may account for the difference. A consistent cloud base is a significant feature of many tropical montane forests2,54, and regular low cloud is assumed to maintain the diversity and abundance of cloud forest epiphytes, and control many of the unique structural and functional features of cloud forests43,44. Indeed, we chose the elevations for this experiment because of a suite of changes in ecosystem structure and function that occur at these elevations, including a step-change in bryophyte and vascular epiphyte biomass40, tree height, above ground biomass, and forest productivity declining42, and soil organic matter increasing41 above 1500 m. Tree diversity also begins to decline above 1500 m in the study region55,56 mimicking the general pattern in the Andes57,58. These clear changes in forest structure, diversity and productivity contrast with smoother changes in climate. Mean temperature, precipitation, and VPD, a measure of moisture stress on plants, all decrease linearly with elevation above 1000 m46.\n\nHigh rainfall in this part of the Andes may mean that epiphytes here are less dependent on cloud immersion to maintain their water balance than their counterparts in other cloud forests. Even in 2005 under drought conditions, total precipitation for the year was 3273 mm. In this pluvial system, cloud base may be less important in determining epiphyte distributions than in other systems. It is noteworthy that the Nadkarni and Solano29 experiment was carried out on the leeward Pacific slope of Monteverde, which is drier than the Caribbean slope59,60. Mean annual precipitation on the Pacific slope is 2155 mm at 1480 m in the cloud forest61, and declines at lower elevations60, and there is a 5-6 month dry season where much of the hydrologic balance is maintained by cloud immersion61. This steep moisture gradient between cloud forest and lower elevations probably leads to a greater dependence of epiphytes on cloud immersion. If this previous study had been carried out on the Caribbean slope, where precipitation is higher at lower elevation60, the results may have been similar to our study. On leeward slopes rainfall compensation may occur, in which epiphyte survival is enhanced by high rainfall even when there is less frequent cloud immersion.\n\nEven though high rainfall may maintain epiphytes under normal conditions in the eastern Andes, droughts do occur, and epiphytes may be adapted to infrequent drought, especially at the lower fringe of the cloud forest. Drought in the Amazon basin during 200562–64 resulted in lower precipitation in the cloud forest. Rainfall at the Peruvian SENAMHI meterological station at Rocotal (13°06′41″S, 71°34′14″, approximately 7 km from the transplant site at 2010 m elevation) for May–August in 2005 was the lowest for any year measured (mean May–August precipitation for 2000–2008: 601 mm; 2005: 175 mm). There was no recorded rainfall in July 2005, the only month during the nine-year measurement period with no recorded precipitation (mean July precipitation: 112 mm). In addition, actual cloud water interception based on fog collectors in place during the experiment did not show a gradient of increasing moisture with elevations during the 2005 dry season (four week total weight of water collected: 1500 m, 1109 g m-2; 1750 m, 35 g m-2; 1900 m, 72 g m-2). All elevations were very dry, and desiccation was evident in bryophytes and non-succulent vascular epiphytes in the study area. However, ramet mortality in undisturbed control mats was not significantly greater than recruitment at any elevation (Figure 1). In addition, mat species composition did not change directionally between years (Figure 3). Thus, undisturbed epiphytes between 1500 and 1800 m in this Andean cloud forest appeared resistant to drought over the one-year time scale of our experiment.\n\nThis resistance to drought may be related to the normally variable and seasonal rainfall at the study site. Annual rainfall totals ranged between 3 and 6 m per year in a five year period not including the 2005 Amazonian drought. Precipitation was lowest in June and July46, when temporary drought is possible, though for no month did PET exceed precipitation at these elevations in most years. However, prolonged (days-to-weeks) periods of direct sun can induce drought stress, and epiphyte species that live in this part of the Andes may possess adaptations for surviving drought, similar to those in lowland dry or seasonal forests36,37,65. Epiphyte drought tolerance is higher in areas where drought occurs more frequently19,66, and many epiphyte species have adaptations for surviving drought – crassulacean acid metabolism, desiccation tolerance, pseudobulbs, succulent leaves and other water-storing organs. Consistent with this idea, Elaphoglossum ramets transplanted from 1500 m had higher survival than those from higher elevations, regardless of the transplant elevation (Figure 2). The stronger response of epiphyte mats transplanted from normally cloud immersed elevations (i.e. 1800 m) compared to those transplanted from lower elevations suggests that lower elevation populations may be better adapted to drought stress due to less frequent cloud immersion. Another example of locally adapted epiphytes was observed in subtropical China, where bryophytes transplanted downslope lost biomass, while in situ measurements showed no change in biomass across the gradient38.\n\n\nConclusion\n\nGreater epiphyte resistance to drought in this part of the Andes compared to previous studies may indicate that even seemingly benign dry seasons or dry periods can be important for structuring epiphyte communities, with potential implications for larger scale patterns of diversity. More generally, while epiphyte response to global climate change on tropical mountains is discussed in the literature18,19,29,38, tropical mountains and their climates are highly heterogeneous, and predictions may defy all but the broadest generalizations. Fundamental differences in the climate and biogeographical contexts may lead to differences in species response to climate change. Long-term experimental studies in tropical montane systems are needed to understand the drivers of patterns of epiphyte abundance, in particular why there is a change in biomass and abundance at putative ‘cloud base’ (which is correlated with changes throughout the ecosystem), and how these diverse communities will respond to climate change. While our experiment suggests that epiphytes in our study system show some resistance to climate change, climate models predict more severe droughts in parts of the Andes20,21. Pervasive changes in the tree canopy of the western Amazon following the 2005 Amazon drought persisted until an even stronger drought in 201067; it is unknown whether similarly long-lasting effects were present in Andean cloud forest. Given the keystone position of epiphytes in cloud forests, drought-induced changes in epiphyte communities could have cascading effects throughout the ecosystem.\n\n\nData availability\n\nFigshare: Survey of epiphytes before and after transplantation across an altitudinal gradient of increasing cloud immersion along the eastern slope of the Andes, Peru, http://dx.doi.org/10.6084/m9.figshare.89497968.", "appendix": "Author contributions\n\n\n\nJMR and MRS conceived and designed the experiments. JMR performed the experiments and analyzed the data. JMR and MRS wrote the manuscript. The authors declare that the experiments described comply with the current laws of Peru.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nFunding was generously provided by a Wake Forest University Biology Department Vecellio Fund grant to J.M. Rapp, and grants from the Gordon and Betty Moore Foundation Andes to Amazon Program, NSF EAR 0711414, NSF DEB-0237684, and an NSF REU supplement DEB-0237684 to M.R. Silman.\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 paper is a product of the Andes Biodiversity and Ecosystem Research Group (ABERG). We would like to thank Norma Salinas for help with permits, the Cock-of-the-Rock Lodge for letting us set up the experiment on their land, Luis Imunda for clearing a trail, Richard Amick for helping with canopy access, and William Farfan for species identification.\n\n\nSupplementary materials\n\nExperimental mats in a tree at 1800 m were all on the same large horizontal branch. From lower left to upper right: two mats transplanted from trees at 1500 and 1650 m, an open space awaiting a transplanted mat from 1800 m, and a surveyed undisturbed control mat.\n\nA transplanted mat was removed from this branch at 1650 m.\n\nA mat from another tree was transplanted into the same location.\n\nA control epiphyte mat after being surveyed in 2005. This mat at 1800 m elevation included primarily Elaphoglossum and Schaphyglottis.\n\n\nReferences\n\nFoster P: The potential negative impacts of global climate change on tropical montane cloud forests. Earth Sci Rev. 2001; 55(1–2): 73–106. Publisher Full Text\n\nLoope LL, Giambelluca TW: Vulnerability of island tropical montane cloud forests to climate change, with special reference to East Maui, Hawaii. Clim Change. 1998; 39(2–3): 503–517. Publisher Full Text\n\nHerzog S, Martinez R, Jorgensen P, et al.: Climate change and biodiversity in the tropical Andes. Inter-American Institute for Global Change Research (IAI) and Scientific Committee on Problems of the Environment (SCOPE) 2011. Reference Source\n\nMyers N, Mittermeier RA, Mittermeier CG, et al.: Biodiversity hotspots for conservation priorities. Nature. 2000; 403(6772): 853–858. Publisher Full Text\n\nGentry AH: Tropical Forest Biodiversity-Distributional Patterns and Their Conservational Significance. Oikos. 1992; 63(1): 19–28. Reference Source\n\nKnapp S: Assessing patterns of plant endemism in neotropical uplands. Bot Rev. 2002; 68(1): 22–37. Publisher Full Text\n\nKuper W, Kreft H, Nieder J, et al.: Large-scale diversity patterns of vascular epiphytes in Neotropical montane rain forests. J Biogeogr. 2004; 31(9): 1477–1487. Publisher Full Text\n\nCruz-Angon A, Baena ML, Greenberg R: The contribution of epiphytes to the abundance and species richness of canopy insects in a Mexican coffee plantation. J Trop Ecol. 2009; 25(5): 453–463. Publisher Full Text\n\nCruz-Angon A, Greenberg R: Are epiphytes important for birds in coffee plantations? An experimental assessment. J Appl Ecol. 2005; 42(1): 150–159. Publisher Full Text\n\nNadkarni NM, Matelson TJ: Bird Use of Epiphyte Resources in Neotropical Trees. Condor. 1989; 91(4): 891–907. Reference Source\n\nYanoviak SP, Nadkarni NM, Solano R: Arthropod assemblages in epiphyte mats of Costa Rican cloud forests. Biotropica. 2007; 39(2): 202–210. Publisher Full Text\n\nGomez-Peralta D, Oberbauer SF, McClain ME, et al.: Rainfall and cloud-water interception in tropical montane forests in the eastern Andes of Central Peru. For Ecol Manage. 2008; 255(3–4): 1315–1325. Publisher Full Text\n\nHolscher D, Kohler L, van Dijk A, et al.: The importance of epiphytes to total rainfall interception by a tropical montane rain forest in Costa Rica. J Hydrol. 2004; 292(1–4): 308–322. Publisher Full Text\n\nMunoz-Villers LE, Holwerda F, Gomez-Cardenas M, et al.: Water balances of old-growth and regenerating montane cloud forests in central Veracruz, Mexico. J Hydrol. 2012; 462–463: 53–66. Publisher Full Text\n\nNadkarni NM, Schaefer D, Matelson TJ, et al.: Biomass and nutrient pools of canopy and terrestrial components in a primary and a secondary montane cloud forest, Costa Rica. For Ecol Manage. 2004; 198(1–3): 223–236. Publisher Full Text\n\nUmana NHN, Wanek W: Large canopy exchange fluxes of inorganic and organic nitrogen and preferential retention of nitrogen by epiphytes in a tropical lowland rainforest. Ecosystems. 2010; 13(3): 367–381. Publisher Full Text\n\nWeathers KC: The importance of cloud and fog in the maintenance of ecosystems. Trends Ecol Evol. 1999; 14(6): 214–215. PubMed Abstract \| Publisher Full Text\n\nBenzing DH: Vulnerabilities of tropical forests to climate change: The significance of resident epiphytes. Clim Change. 1998; 39(2–3): 519–540. Publisher Full Text\n\nZotz G, Bader MY: Epiphytic Plants in a Changing World: Global Change Effects on Vacular and Non-vascular Epiphytes. in Progress in Botany, U. Lüttge, et al., Editors. Springer-Verlag: Berlin Heidelberg. 2009; 70. : 147–170. Publisher Full Text\n\nNeelin JD, Munnich M, Su H, et al.: Tropical drying trends in global warming models and observations. Proc Nat Acd Sci. 2006; 103(16): 6110–6115. Publisher Full Text\n\nUrrutia R, Vuille M: Climate change projections for the tropical Andes using a regional climate model: Temperature and precipitation simulations for the end of the 21st century. J Geophys Res Atmos. 2009; 114(D2). 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\n\nMartinez R, Ruiz D, Andrade M, et al.: Synthesis of the climate of the tropical Andes, in Climate change and biodiversity in the tropical Andes. SK Herzog, et al., Editors. Inter-American Institute for Global Change Research (IAI) and Scientific Committee on Problems of the Environment (SCOPE) 2011. Reference Source\n\nCowling SA, Shin Y, Pinto E, et al.: Water recycling by Amazonian vegetation: coupled versus uncoupled vegetation-climate interactions. Philos Trans R Soc Lond B Biol Sci. 2008; 363(1498): 1865–1871. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPinto E, Shin Y, Cowling SA, et al.: Past, present and future vegetation-cloud feedbacks in the Amazon Basin. Clim Dynamics. 2009; 32(6): 741–751. Publisher Full Text\n\nStill CJ, Foster PN, Schneider SH: Simulating the effects of climate change on tropical montane cloud forests. Nature. 1999; 398(6728): 608–610. Publisher Full Text\n\nPounds JA, Fogden MPL, Campbell JH: Biological response to climate change on a tropical mountain. Nature. 1999; 398(6728): 611–615. Publisher Full Text\n\nAnchukaitis KJ, Evans MN: Tropical cloud forest climate variability and the demise of the Monteverde golden toad. Proc Natl Acad Sci U S A. 2010; 107(11): 5036–5040. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNadkarni NM, Solano R: Potential effects of climate change on canopy communities in a tropical cloud forest: an experimental approach. Oecologia. 2002; 131(4): 580–586. Publisher Full Text\n\nHamilton LS, Juvik JO, Scatena FN: The Puerto Rico Tropical Cloud Forest Symposium: Introduction and Workshop Synthesis, in Tropical Montane Cloud Forests. LS Hamilton, JO Juvik, and FN Scatena, Editors. Springer-Verlag: New York 1994. Reference Source\n\nStadtmuller T: Cloud Forests in the Humid Tropics: A Bibliographic Review. Centro Agronómico Tropical de Investigación y Enseñanza: Turrialba, Costa Rica: United Nations University Press: Tokyo. 1987; 82. Reference Source\n\nBruijnzeel LA: Hydrological functions of tropical forests: not seeing the soil for the trees? Agric Ecosystems Environ. 2004; 104(1): 185–228. Publisher Full Text\n\nLawton RO, Nair US, Pielke Sr RA, et al.: Climatic impact of tropical lowland deforestation on nearby montane cloud forests. Science. 2001; 294(5542): 584–587. PubMed Abstract \| Publisher Full Text\n\nPounds JA, Bustamante MR, Coloma LA, et al.: Widespread amphibian extinctions from epidemic disease driven by global warming. Nature. 2006; 439(7073): 161–167. PubMed Abstract \| Publisher Full Text\n\nKilleen TJ, Douglas M, Consiglio T, et al.: Dry spots and wet spots in the Andean hotspot. J Biogeogr. 2007; 34(8): 1357–1373. Publisher Full Text\n\nAndrade JL: Dew deposition on epiphytic bromeliad leaves: an important event in a Mexican tropical dry deciduous forest. J Trop Ecol. 2003; 19(5): 479–488. Publisher Full Text\n\nBader MY, Menke G, Zotz G: Pronounced drought tolerance characterizes the early life stages of the epiphytic bromeliad Tillandsia flexuosa. Funct Ecol. 2009; 23(3): 472–479. Publisher Full Text\n\nSong L, Liu WY, Nadkarni NM: Response of non-vascular epiphytes to simulated climate change in a montane moist evergreen broad-leaved forest in southwest China. Biol Conserv. 2012; 152: 127–135. Publisher Full Text\n\nMalhi Y, Silman M, Salinas N, et al.: Introduction: Elevation gradients in the tropics: laboratories for ecosystem ecology and global change research. Glob Chang Biol. 2010; 16(12): 3171–3175. Publisher Full Text\n\nHorwath AB: Epiphytic Bryophytes as Cloud Forest Indicators: Stable Isotopes, Biomass and Diversity along an Altitudinal Gradient in Peru. Ph.D. Thesis, University of Cambridge: Cambridge, U.K. 2011; 1–284.\n\nZimmermann M, Meir P, Bird MI, et al.: Climate dependence of heterotrophic soil respiration from a soil-translocation experiment along a 3000 m tropical forest altitudinal gradient. Eur J Soil Sci. 2009; 60(6): 895–906. Publisher Full Text\n\nGirardin CAJ, Malhi Y, Aragao LEOC, et al.: Net primary productivity allocation and cycling of carbon along a tropical forest elevational transect in the Peruvian Andes. Glob Chang Biol. 2010; 16(12): 3176–3192. Publisher Full Text\n\nFrahm JP, Gradstein SR: An Altitudinal Zonation of Tropical Rain-Forests Using Byrophytes. J Biogeogr. 1991; 18(6): 669–678. Publisher Full Text\n\nGrubb PJ: Factors controlling the distribution of forest types on tropical mountains: new facts and a new perspective, in Altitudinal zonationin Malesia, JR Flenley, Editor, University of Hull, Department of Geography: Hull, England. 1974; 13–46.\n\nZimmermann M, Meir P, Bird MI, et al.: Temporal variation and climate dependence of soil respiration and its components along a 3000 m altitudinal tropical forest gradient. Global Biogeochem Cycles. 2010; 24(4). Publisher Full Text\n\nRapp JM, Silman MR: Diurnal, seasonal, and altitudinal trends in microclimate across a tropical montane cloud forest. Clim Res. 2012; 55(1): 17–32. Publisher Full Text\n\nMotzer T, Munz N, Kuppers M, et al.: Stomatal conductance, transpiration and sap flow of tropical montane rain forest trees in the southern Ecuadorian Andes. Tree Physiol. 2005; 25(10): 1283–1293. PubMed Abstract \| Publisher Full Text\n\nCunningham SC: Stomatal sensitivity to vapour pressure deficit of temperate and tropical evergreen rainforest trees of Australia. Trees. 2004; 18(4): 399–407. Publisher Full Text\n\nDial R, Tobin SC: Description of arborist methods for forest canopy access and movement. Selbyana. 1994; 15(2): 24–37. Reference Source\n\nVenables WN, Ripley BD: Modern Applied Statistics with S. 4th ed New York: Springer 2002: 497. Reference Source\n\nBolker BM, Brooks ME, Clark CJ, et al.: Generalized linear mixed models: a practical guide for ecology and evolution. Trends Ecol Evol. 2009; 24(3): 127–135. PubMed Abstract \| Publisher Full Text\n\nR Development Core Team, R: A language and environment for statistical computing. R Foundation for Statistical Computing: Vienna, Austria 2012. Reference Source\n\nOksanen J, Blanchet FG, Kindt R, et al.: Vegan: Community Ecology Package. 2010. Reference Source\n\nRitter A, Regalado CM, Aschan G: Fog reduces transpiration in tree species of the Canarian relict heath-laurel cloud forest (Garajonay National Park, Spain). Tree Physiol. 2009; 29(4): 517–528. PubMed Abstract \| Publisher Full Text\n\nFierer N, McCain CM, Meir P, et al.: Microbes do not follow the elevational diversity patterns of plants and animals. Ecology. 2011; 92(4): 797–804. PubMed Abstract \| Publisher Full Text\n\nJankowski JE, Merkord CL, Rios WF, et al.: The relationship of tropical bird communities to tree species composition and vegetation structure along an Andean elevational gradient. J Biogeogr. 2013; 40(5): 950–962. Publisher Full Text\n\nGentry AH: Patterns of Diversity and Floristic Composition in Neotropical Montane Forests, in Biodiversity and Conservation of Neotropical Montane Forests. SP Churchill, et al., Editors. The New York Botanical Garden: New York. 1995; p.103–126. Reference Source\n\nGentry AH: Changes in plant community diversity and floristic composition on environmental and geographical gradients. Ann Mo Bot Gard. 1988; 75(1): 1–34. Reference Source\n\nGuswa AJ, Rhodes AL, Newell SE: Importance of orographic precipitation to the water resources of Monteverde, Costa Rica. Adv Water Resour. 2007; 30(10): 2098–2112. Publisher Full Text\n\nHäger A, Dohrenbusch A: Hydrometeorology and structure of tropical montane cloud forests under contrasting biophysical conditions in north-western Costa Rica. Hydrol Process. 2011; 25(3): 392–401. Publisher Full Text\n\nClark KL, Lawton RO, Butler PR: The physical environment, in Monteverde: Ecology and conservation of a tropical cloud forest. NM Nadkarni and NT Wheelwright, Editors. Oxford University Press: New York. 2000; 15–38. Reference Source\n\nMarengo JA, Nobre CA, Tomasella J, et al.: The drought of Amazonia in 2005. J Climate. 2008; 21(3): 495–516. Publisher Full Text\n\nMarengo JA, Nobre CA, Tomasella J, et al.: Hydro-climate and ecological behaviour of the drought of Amazonia in 2005. Philos Trans R Soc Lond B Biol Sci. 2008; 363(1498): 1773–1778. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZeng N, Yoon JH, Marengo JA, et al.: Causes and impacts of the 2005 Amazon drought. Environ Res Lett. 2008; 3(1): 1–9. Publisher Full Text\n\nStancato GC, Mazzafera P, Buckeridge MS: Effect of a drought period on the mobilisation of non-structural carbohydrates, photosynthetic efficiency and water status in an epiphytic orchid. Plant Physiol Biochem. 2001; 39(11): 1009–1016. Publisher Full Text\n\nHiguera D, Wolf JHD: Vascular epiphytes in dry oak forests show resilience to anthropogenic disturbance, Cordillera Oriental, Colombia. Caldasia. 2010; 32(1): 161–174. Reference Source\n\nSaatchi S, Asefi-Najafabady S, Malhi Y, et al.: Persistent effects of a severe drought on Amazonian forest canopy. Proc Natl Acad Sci U S A. 2013; 110(2): 565–570. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRapp JM, Silman MR: Survey of epiphytes before and after transplantation across an altitudinal gradient of increasing cloud immersion along the eastern slope of the Andes, Peru. figshare, 2013. Data Source" }	[ { "id": "3588", "date": "10 Mar 2014", "name": "Peter Hietz", "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 addresses the question how epiphyte communities in tropical montane forests may be affected by climate change by transplanting epiphyte mats between forests at different altitudes. This is a reasonable proxy of what might happen under climate change with moderate temperature changes and substantial changes in cloud base height and thus available humidity. The paper is well written, the experiment is clearly described and the analysis is logical (if complex). I do suggest some additional thoughts and caveats - for the discussion mainly. As one might expect, the outcome of such a study (four genera, three altitudes, three parameters evaluated) is complex, with significant and (mostly) non-significant effects. Overall the effect was perhaps less than expected (at least compared to a similar study by Nadkarni & Solano (2009) from Costa Rica). In this case, one tends to pick the significant effects that confirm expectations and ignore those that do not. The authors should avoid simplifying too much, particularly in the abstract, which is what most people will read. The abstract states that \"Ramet mortality increased, recruitment decreased, and population size declined for epiphytes in mats transplanted down slope from the highest elevation\". This appears to contrast with the results (\"Across all species, there were no significant effects on survival of any of the treatments for mats transplanted across elevations\"). Details of the experimental species used are covered in the methods and results sections, however it seems that the plants were either identified only to the genus level or several species (identified or not) were included within each genus. Unless the authors are sure that only one species per genus was used, they should refer to the groups as \"genus\", and not \"species\". It is a substantial difference if we compare one species with individuals adapted to different climates within their phenotypic plasticity, or species differing in their evolutionary adaptations. This should be made clear in the methods section and acknowledged in the discussion. Under drought, plants or ramets may completely die off, but drought first results in weaker plants that exhaust their resources. As such, the effect of being transplanted to a different climate may take more than one year to produce significant effects on population size or also on other parameters studied. This should be acknowledged. The discussion suggests that epiphytes may be fairly drought tolerant in the forest studied because they must have survived unusually dry years (such as 2005, at the beginning of the experiment). This is a good point, and to put this into context it would be good to show precipitation trends (perhaps monthly sums for the closest station available) over several years preceding and including the experiment. Minor comments:Please explain how the daily photographs used to compare the cloud base height were taken and analyzed.The correct name is Scaphyglottis, not Schaphyglottis.Why does Table 4 have three elevations for Elaphoglossum, one for Maxillaria (all not significant) and none for the other two genera?”", "responses": [ { "c_id": "834", "date": "23 May 2014", "name": "Joshua Rapp", "role": "Author Response", "response": "We thank Dr. Hietz for taking time to review our paper and for the constructive comments. These have improved the paper. We respond to these comments below and describe the changes made to the paper in response to these comments. Dr. Hietz’ original comments are in [brackets].[As one might expect, the outcome of such a study (four genera, three altitudes, three parameters evaluated) is complex, with significant and (mostly) non-significant effects. Overall the effect was perhaps less than expected (at least compared to a similar study by Nadkarni & Solano (2009) from Costa Rica). In this case, one tends to pick the significant effects that confirm expectations and ignore those that do not. The authors should avoid simplifying too much, particularly in the abstract, which is what most people will read. The abstract states that \"Ramet mortality increased, recruitment decreased, and population size declined for epiphytes in mats transplanted down slope from the highest elevation\". This appears to contrast with the results (\"Across all species, there were no significant effects on survival of any of the treatments for mats transplanted across elevations\"). ]We have revised the text to be more precise regarding the observed results, especially in the Discussion. In the Abstract we now state that effects were stronger and more negative for epiphytes moved from the highest elevation as compared to those transplanted from other elevations. While this is certainly still a simplification of the results, it does accurately represent the results since significant effects were largely confined to mats transplanted down-slope. [Details of the experimental species used are covered in the methods and results sections, however it seems that the plants were either identified only to the genus level or several species (identified or not) were included within each genus. Unless the authors are sure that only one species per genus was used, they should refer to the groups as \"genus\", and not \"species\". It is a substantial difference if we compare one species with individuals adapted to different climates within their phenotypic plasticity, or species differing in their evolutionary adaptations. This should be made clear in the methods section and acknowledged in the discussion. ]We identified each taxa to morpho-species. This is stated in the first paragraph of the Data collection section of the methods, but was obviously not clear enough. We used species after this point for brevity, but to make it clear to the reader, we have now changed ‘species’ to ‘morpho-species’ whenever referring directly to the taxa in the study. We have left ‘species’ in place where it is used in a more general discussion (i.e. when discussing species composition, species ranges, etc.). We were unable to identify taxa to species definitively because most species were not reproductive during the survey periods. Therefore, while we are fairly certain that our morpho-species were single biological species, it is possible that one or more contain multiple cryptic species. We describe in more detail in the Data collection section why morpho-species were used, and add a sentence acknowledging that taxonomic uncertainty could have affected our results. [Under drought, plants or ramets may completely die off, but drought first results in weaker plants that exhaust their resources. As such, the effect of being transplanted to a different climate may take more than one year to produce significant effects on population size or also on other parameters studied. This should be acknowledged. ]Very good point. We added a discussion of this idea to the second paragraph of the Discussion (this is a new paragraph). [The discussion suggests that epiphytes may be fairly drought tolerant in the forest studied because they must have survived unusually dry years (such as 2005, at the beginning of the experiment). This is a good point, and to put this into context it would be good to show precipitation trends (perhaps monthly sums for the closest station available) over several years preceding and including the experiment.] Added figure (Figure 4) of monthly rainfall at the closest climate station (Rocotal, 2010m elevation) to the Discussion.Minor comments:[Please explain how the daily photographs used to compare the cloud base height were taken and analyzed.]Now explained in the Methods.[The correct name is Scaphyglottis, not Schaphyglottis.]Thank you. Fixed throughout.[Why does Table 4 have three elevations for Elaphoglossum, one for Maxillaria (all not significant) and none for the other two genera?”]Table 2 shows the abundance of each morpho-species in the initial plots at each elevation. For the ‘within-species’ analysis, we only included source elevations where the morpho-species was abundant (bold in Table 2). Our criteria for “abundant species” is now given in the Methods." } ] }, { "id": "3105", "date": "11 Mar 2014", "name": "Liang 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\nThe purpose of this manuscript is to determine the potential impacts of drought and experimental warming on the performance of vascular epiphytes in a tropical montane cloud forest. There is a lack of information about the effects of climate change on epiphytes in mountain environments. The use of empirical rather than observational data as well as exchanging space for time to demonstrate those effects is also very commendable since such data is extremely rare. Most studies of the effects of a changing climate on plants in mountain environments focus on changes to the tree line or tree species. However, the choice of epiphytes as experimental subjects in this study is also important, since in my opinion, they would be one of the first groups of organisms to reflect the effects of climate change. Epiphytes have a more direct link to atmospheric conditions and changes to those conditions than plants that root in the soil, and thus are more sensitive to moisture levels as noted by the authors. This study clearly creates unique data on the impact of climate change on epiphytes and demonstrates an alternative, highly sensitive method of detecting changes in atmospheric conditions in mountain environments. The article is well constructed and clear and the statistical analyses appropriate. Considering that this experiment was conducted in the canopy, the authors must have made a great effort to finish it. Although the underpinnings of this work are scientifically sound, and some results correlate well with data previously reported for vascular epiphytes by other authors (Nadkarni and Solano, 2002), there are several major problems that I feel should be addressed:The time span of the experiment is only one year, which is too short. Microclimatic data was not monitored in the three studied elevations during the experimental period - which may obscure the interpretation of the results. There should be more species with different functional types, which could support the third hypothesis that different functional types (e.g. strap-leaf ferns, orchids, ericaceous shrubs) do not respond similarly to moisture gradients. If it is not possible for you to find enough species I would suggest not using ‘functional types’. I have provided a more detailed critique below: Title:The title is well chosen although I would suggest adding the location where the study was conducted (e.g. “in the Andes cloud forest”). As noted by the authors, “Cloud forests vary worldwide, with differences in cloud base height and the proportion of moisture received by the vegetation via cloud stripping versus rainfall” and “High rainfall in this part of the Andes may mean that epiphytes here are less dependent on cloud immersion to maintain their water balance than their counterparts in other cloud forests.” - This is not a general study concerned about all epiphytes, but a local study restricted to four typical vascular epiphytes in the Andes cloud forest.Abstract:The abstract is basically well written, but the summary of the main findings do not seem appropriate. In the paper, the authors summarize that “Ramet mortality increased, recruitment decreased, and population size declined for epiphytes in mats transplanted down slope from the highest elevation, into warmer and drier conditions”, which is not strongly supported by the results. In the results, the authors mention that “Across all species, there were no significant effects on survival of any of the treatments for mats transplanted across elevations (Table 5 and Figure 2). For recruitment and population change, there was a significant interaction between source and transplant elevation (Table 5), with both positively associated with elevation for mats transplanted from 1500 and 1800 m, but negatively associated with altitude for mats from 1650 m (Figure 2).” Different epiphyte species and different parameters showed different response to treatments and sources. The results are much more complicated than what you summarized. The authors mentioned that their reciprocal transplant design allowed them to distinguish between the effects of moving mats away from their home elevation versus moving plants into lower moisture conditions. As a key control, results concerned about the transplant effect should be included in the abstract. The authors mention biomass loss in the last sentence, but this conclusion was based on other studies (I guess it is based on Song’s results), not their own. Introduction:Paragraph 1: I suggest adding a sentence to illuminate why epiphytes are sensitive to water availability and temperature. Paragraph 4: The authors mention “Do epiphyte species of different functional types (e.g. strap-leaf ferns, orchids, ericaceous shrubs) respond similarly to moisture gradients?” as one of the objectives of the study. If so, why were only four species studied? There is only one fern, one shrub, and two orchids in the experiment. It is not possible, in my opinion, to draw any useful conclusions of functional types with such a limited selection. Paragraph 5: this paragraph is repeating much of the fourth paragraph, I suggest combining the two. Materials and Methods:The authors could provide a more detailed description of the four target genera, e.g., “What are the elevational ranges of the target genera in the Andes?” “What is the abundance of the four target genera at different elevations (1500m, 1650m, 1800m)?” etc. This information would be valuable to the interpretation of the results. Have the authors monitored the microclimatic data of the studied elevations during the experimental period? The details of this data are critical to the interpretation of the results yet it has not been included. Although the July climate for weather stations (Table 1) showed the general pattern of temperature and moisture level along the experimental altitudes it is not the microclimatic data of the study sites, especially as no data was provided at an elevation of 1650m. The time span of the experiment is only one year. Longer term monitoring (at least 2 years) at regular intervals (e.g. one year) would be much more valuable. In addition, I would suggest that the authors monitor other physiological parameters such as photosynthesis and chlorophyll fluorescence in their future work. Their conclusions would be much stronger based on these data. Discussion:The title of the paper is “Epiphyte response to drought and experimental warming”; however, the authors do not mention warming in the discussion section. The authors should focus the discussion section more on how different epiphytes respond to drought and warming in this study rather than the reasons for the relative resistance of epiphytes to moisture stress. The authors overstate their results in some places. For example, they mentioned “Vascular epiphytes transplanted down slope from our highest elevation had lower demographic performance when transplanted to the lowest elevation” in the first sentence of the discussion section. This is only partly true for population change, while not true for survival and recruitment (Fig. 2). Although functional types are mentioned, the authors only have 4 species belonging to 3 functional types. Paragraph 1: “In general, it appears that epiphytes responded to water stress but we also found evidence for local adaptation.” Both temperature and water availability changes along the elevations. Why are you sure that epiphytes responded to water stress not temperature? Paragraph 2 and 3: Rainfall compensation and drought tolerance seems self-contradictory. PET: You should provide the full name when it appeared for the first time.", "responses": [ { "c_id": "833", "date": "23 May 2014", "name": "Joshua Rapp", "role": "Author Response", "response": "We thank Dr. Song for the reviewing our paper and providing constructive feedback. Below we respond to these comments and describe the changes made to the paper in response. Dr. Song’s comments are in [brackets].[The time span of the experiment is only one year, which is too short.]While we agree that it would have been ideal to extend the study another year, this was not possible due to logistical constraints. Since the experiment was conducted in 2005-2006, it is not possible to add more time to the experiment. We now discuss the effect of experiment duration in the second paragraph of the Discussion. [Microclimatic data was not monitored in the three studied elevations during the experimental period - which may obscure the interpretation of the results.]We agree that microclimate data during the experiment would have been ideal. Sensors placed at the study site during the experiment stopped recording data soon after deployment. We therefore used data from another period. We added a sentence to the Discussion noting that microclimate data was not available during the period, but refer to rainfall data from the nearby Rocotal climate station. [There should be more species with different functional types, which could support the third hypothesis that different functional types (e.g. strap-leaf ferns, orchids, ericaceous shrubs) do not respond similarly to moisture gradients. If it is not possible for you to find enough species I would suggest not using ‘functional types’.]Since the experimental design focused on plots rather than individual species, we did not have control over the sample size for individual species. In this analysis we used the most abundant species in the plots, to maximize the statistical power to detect differences among treatments. We agree that this experimental design does not allow for a robust test of the hypothesis that functional types may differ in their response to moisture gradients. We therefore removed this hypothesis from the Introduction. However, we include the point in the Discussion, since the experiment does shed some light on differences among functional types, and we feel the patterns are worth mentioning. Title:[The title is well chosen although I would suggest adding the location where the study was conducted (e.g. “in the Andes cloud forest”). As noted by the authors, “Cloud forests vary worldwide, with differences in cloud base height and the proportion of moisture received by the vegetation via cloud stripping versus rainfall” and “High rainfall in this part of the Andes may mean that epiphytes here are less dependent on cloud immersion to maintain their water balance than their counterparts in other cloud forests.” - This is not a general study concerned about all epiphytes, but a local study restricted to four typical vascular epiphytes in the Andes cloud forest.]Title changed to “Epiphyte response to drought and experimental warming in an Andean cloud forest”Abstract:[The abstract is basically well written, but the summary of the main findings do not seem appropriate. In the paper, the authors summarize that “Ramet mortality increased, recruitment decreased, and population size declined for epiphytes in mats transplanted down slope from the highest elevation, into warmer and drier conditions”, which is not strongly supported by the results. In the results, the authors mention that “Across all species, there were no significant effects on survival of any of the treatments for mats transplanted across elevations (Table 5 and Figure 2). For recruitment and population change, there was a significant interaction between source and transplant elevation (Table 5), with both positively associated with elevation for mats transplanted from 1500 and 1800 m, but negatively associated with altitude for mats from 1650 m (Figure 2).” Different epiphyte species and different parameters showed different response to treatments and sources. The results are much more complicated than what you summarized.]We’ve changed to state that epiphytes moved down slope from the highest elevation had stronger and more negative response than epiphytes moved from other elevations. This also addresses the next comment.[The authors mentioned that their reciprocal transplant design allowed them to distinguish between the effects of moving mats away from their home elevation versus moving plants into lower moisture conditions. As a key control, results concerned about the transplant effect should be included in the abstract. The authors mention biomass loss in the last sentence, but this conclusion was based on other studies (I guess it is based on Song’s results), not their own.]We revised the last sentence to focus on how the environmental and evolutionary context may affect epiphyte response to climate change.Introduction:[Paragraph 1: I suggest adding a sentence to illuminate why epiphytes are sensitive to water availability and temperature.]We added a clause to the sentence: “Cloud immersion is important for many epiphyte species to maintain a positive water balance and avoid desiccation” making it clear that: “this makes them sensitive to changes in moisture regimes.” In the next paragraph we describe how cloud formation and temperature are linked, and do not feel that it would add any clarity to the manuscript to repeat that here.[Paragraph 4: The authors mention “Do epiphyte species of different functional types (e.g. strap-leaf ferns, orchids, ericaceous shrubs) respond similarly to moisture gradients?” as one of the objectives of the study. If so, why were only four species studied? There is only one fern, one shrub, and two orchids in the experiment. It is not possible, in my opinion, to draw any useful conclusions of functional types with such a limited selection.]We deleted Question (3).[Paragraph 5: this paragraph is repeating much of the fourth paragraph, I suggest combining the two.]We deleted the last two sentences referring to functional types. The rest of the paragraph elaborates on the approach outlined in the preceding paragraph. We feel that for clarity it is useful to first describe the research questions generally, and then operationalize the general questions. For this reason, we first describe how we expect performance to change with treatments, and then, in the next paragraph, describe the specific performance metrics used, and more specific predictions of how we expected them to change. We have therefore left the structure as is, with one general and one more specific paragraph. Materials and Methods:[The authors could provide a more detailed description of the four target genera, e.g., “What are the elevational ranges of the target genera in the Andes?” “What is the abundance of the four target genera at different elevations (1500m, 1650m, 1800m)?” etc. This information would be valuable to the interpretation of the results.]All of the genera have broad ranges in Peru, encompassing the study elevations. However, this information is less useful than the elevational distributions of the morpho-species sampled. We have added text in a new section (Focal species) of the Methods to point out more clearly the differences in abundance with elevation of the different target taxa. Table 2 gives the abundance of each the target morpho-species at each elevation.[Have the authors monitored the microclimatic data of the studied elevations during the experimental period? The details of this data are critical to the interpretation of the results yet it has not been included. Although the July climate for weather stations (Table 1) showed the general pattern of temperature and moisture level along the experimental altitudes it is not the microclimatic data of the study sites, especially as no data was provided at an elevation of 1650m.]Unfortunately, temperature and humidity sensors at the study site during the experiment stopped recording data soon after deployment, but this was not discovered until the end of the experiment. Therefore, microclimate data during the experiment is not available. Here we are reporting data for sensors deployed for a different study begun after the experiment that is the focus of this paper. We think this data is still useful, given that it is close to the experimental locations, and better than not reporting any climate data.[The time span of the experiment is only one year. Longer term monitoring (at least 2 years) at regular intervals (e.g. one year) would be much more valuable. In addition, I would suggest that the authors monitor other physiological parameters such as photosynthesis and chlorophyll fluorescence in their future work. Their conclusions would be much stronger based on these data.] While we agree that longer term monitoring would be ideal, this was not possible given logistical constraints. Here we are reporting what we did. We have added a sentence to the Discussion (2nd paragraph) stating that physiological measurements would be helpful.Discussion:[The title of the paper is “Epiphyte response to drought and experimental warming”; however, the authors do not mention warming in the discussion section. The authors should focus the discussion section more on how different epiphytes respond to drought and warming in this study rather than the reasons for the relative resistance of epiphytes to moisture stress.]We now mention temperature and moisture together in the first part of the Discussion, since it is not possible from our experimental design to separate the effects of each.[The authors overstate their results in some places. For example, they mentioned “Vascular epiphytes transplanted down slope from our highest elevation had lower demographic performance when transplanted to the lowest elevation” in the first sentence of the discussion section. This is only partly true for population change, while not true for survival and recruitment (Fig. 2). Although functional types are mentioned, the authors only have 4 species belonging to 3 functional types. ]This is a good point. We have changed the text to place less emphasis on functional types. We still mention differences in functional types as a potential factor in the species-level differences observed, but also include the caveat that our experimental design does not allow the rigorous testing of this hypothesis.[Paragraph 1: “In general, it appears that epiphytes responded to water stress but we also found evidence for local adaptation.” Both temperature and water availability changes along the elevations. Why are you sure that epiphytes responded to water stress not temperature?]Good point. See above.[Paragraph 2 and 3: Rainfall compensation and drought tolerance seems self-contradictory.]These two effects do act in opposite directions, but both are possible, and non-mutually exclusive. Rainfall compensation could be the usual rule, with occasional drought also providing a selective pressure. We now describe this more clearly at the end of section 1 of the Discussion.[PET: You should provide the full name when it appeared for the first time.]Thank you for catching this. Changed to potential evapotranspiration. We also deleted PET, since it only occurs once." } ] } ]	1	https://f1000research.com/articles/3-7
https://f1000research.com/articles/3-122/v1	06 Jun 14	{ "type": "Research Article", "title": "Tooth eruption sequence and dental crowding: a case-control study", "authors": [ "Vahid Moshkelgosha", "Negar Khosravifard", "Ali Golkari", "Vahid Moshkelgosha", "Negar Khosravifard" ], "abstract": "When cases of dental crowding are identified and diagnosed promptly, interceptive orthodontics is particularly successful.Aim: To assess the differences in the eruption sequence of the mandibular canine and first premolar teeth in children with and without dental crowding.Materials and Methods: Children who attended the Shiraz Dental School's orthodontic clinic (Iran) from September to December 2012 were enrolled in this case-control study. Tooth size arch length discrepancy (TSALD) of all 8-10 year olds was calculated from patients’ dental models. Thirty-six children were randomly selected from those with TSALD of equal or less than 4mm (those with crowding). Each selected case was matched for sex and age with another child (as control) with TSALD>−4mm attending the same clinic, in the same time period. The existing panoramic radiographs were traced and the eruption percentages were measured for mandibular canine and first premolar teeth. The mean difference between canine and first premolar eruption percentages was compared between the case and control groups using the SPSS (version PASW 18) software and a paired sample t-test.Results: Canine and first premolar eruption percentages in the case group were 65.82±13.00 and 78.92±10.15 percent, respectively. The mean eruption percentages for canines and first premolars of the control group were 74.12±14.55 and 75.47±11.60 percent, respectively. There was a significant difference in pre-eruptive positions of canine and first premolar teeth in those with moderate to severe crowding when compared to the control group (p<0.001).Conclusion: These findings may improve the early diagnosis of children with high risk of developing moderate to severe crowding during mixed dentition.", "keywords": [ "Interceptive orthodontics", "tooth eruption", "tooth crowding" ], "content": "Introduction\n\nThe National Health and Nutrition Examination Survey (NHANES III, 1998) reported dental crowding as the most prevalent form of malocclusion among children in the United States, with about 50% having some degrees of crowding in the mixed dentition that worsened as they stepped into adolescence and adulthood1.\n\nDespite the frequent advancements in treatment modalities and the use of high technology equipment in contemporary orthodontics, little attempt has been made to advance preventive orthodontic services. Prevention and interception of orthodontic problems are major concerns as they can improve the quality of life of people and save their money and time2,3. Preventing a developing malocclusion or intercepting its path is always more economic and less complicated than correcting the resulting malocclusion later4,5. In many countries, due to the shortage of specialists or the inability of the society to afford treatment, delivering orthodontic treatment after crowding has developed is not possible. However, preventive services are much cheaper and can be easily delivered by general dental practitioners6. Although interceptive treatment techniques are simple in nature, a sound diagnosis is essential. Therefore, the ability to predict future crowding in a child is vital.\n\nOn the other hand, performing interceptive orthodontic procedures (such as serial extraction) at the right time is very important. The appropriate age for most interceptive interventions is when children are in the mixed dentition phase7. To predict severe crowding in a child, clinicians often use some diagnostic clues such as the premature exfoliation of primary canines, prominent bulging in the canine area and the crescent area of root resorption in roots of primary canines8. It is also proposed that the variations in teeth eruption may be an important aspect of crowding9. Although a few studies have supposed averages or standards for tooth eruption sequence9,10, little has been done to understand the relationship between tooth eruption sequence variations and dental crowding.\n\nSampson and Richards, for instance, tested the hypothesis that pre-eruptive tooth positions might forecast crowding and proposed that a buccal eruption path of a mandibular canine indicates an insufficient space in the dental arch11. Moorrees and Reed found that the utilization of leeway space depends on the sequence of shedding and eruption of the mandibular teeth12. In another study, a low but significant correlation was found between increased crowding in the mandibular segment and the retardation of early phases of canine eruption13. More recently, Lange has claimed that those children whose teeth eruption pattern does not follow the standard sequence are at greater risk of developing crowding14. His study has been conducted on a confined population, therefore, it warrants further investigations to see whether the same results can be observed in other ethnics.\n\nConsidering the shortcomings of the literature in correlating the tooth eruption sequence to the possible lack of space, this case-control study was designed to better understand whether the mandibular teeth eruption sequence differs in patients with moderate to severe dental arch crowding.\n\n\nMaterials and methods\n\nA case-control study was designed and approved by the Shiraz University of Medical Sciences’ Orthodontic Research Center in Shiraz, Iran (Approval number 89-01-37-1940[8716]). Children aged from 8 to 10 years that were admitted to the Shiraz Dental School for orthodontic treatment during September to December of 2012 were enrolled in the study. Patients with a history of metabolic disease, nutrition deficiency, traumatic accidents to jaws and premature tooth loss as well as patients with missing data (broken plaster models, partially erupted anterior teeth and poor quality radiographs) were excluded from the study. The objectives and process of the study were explained to the parents. They were assured of the confidentiality of their and their children’s personal information. They were also assured that participation in this study (or their refusal) had no effect on their course of treatment. Parents were then asked to fill and sign the study’s informed consent form. Children of those who did not give their written consent were also removed from the study’s register. The final list of eligible patients consisted of 327 children. The next step was to divide the eligible children into two groups with and without crowding and to randomly select 36 children from each group.\n\nChildren’s plaster models, which were previously obtained for the purpose of orthodontic treatment, were assessed in a random order to calculate the tooth-size arch size discrepancy (TSALD). To do so, the space required for eruption of ten permanent anterior and premolar teeth was deducted from the space available from molar to molar teeth. The available space was measured by dividing the dental arch into four separate segments (Figure 1) and summing them up, as previously described15. The widths of mandibular incisors teeth, which were already erupted, were measured using a digital caliper with accuracy of 0.01 millimeter and were summed up. The width of un-erupted canine and premolar teeth was estimated using the Moyers table16 and the Tanaka-Johnston formulae17. The average of values gained from the two methods of estimation was used in the study.\n\nThe children were then divided into two groups based on their TSALD: those with a TSALD of equal or less than -4mm (crowding group), and those with a TSLAD of greater than -4mm (no crowding group). A computer generated random sample of 36 children was selected from the \"crowding\" group, as cases (TSLD≤-4mm). For each case, one subject matching sex and age was assigned as a control. The control assignment was conducted using a computer based randomization among the \"no crowding\" group (TSLD>-4mm). Therefore, data from a total of 72 children were used in the final analysis.\n\nThe final sample’s panoramic radiographies, which were previously taken for orthodontic treatment purposes, were traced to determine the eruption percent of mandibular canines and first premolar teeth. The tracing was made on acetate paper with a 3H pencil. As the eruption sequence is symmetrical in the left and right sides, tracing of only one side of the mandible was sufficient18. A line was passed through the cusp tip of un-erupted permanent canine and the center of its predecessor primary canine (Figure 2). The distance from the inferior border of the mandible to the cusp tip of un-erupted permanent canine was measured (a, Figure 2). Also, the distance from the cusp tip of un-erupted permanent canine to the line passing through the cusp tip of the primary molars and permanent first molar was measured (b, Figure 2). The eruption percent of canines was then calculated by dividing the first measurement by sum of both (a/a+b). A similar procedure was done to calculate the eruption percentage of permanent first premolar. The above-mentioned technique to calculate the eruption percentage of un-erupted teeth was adapted from Shumakher and El Hadary19.\n\nAll measurements were performed twice by two experienced dentists (V.M. and N.K.). SPSS (version PASW 18) software was used for data entry and analysis. The Pearson correlation test was used to evaluate the reliability of measurements reported by the two assessors (Table 1). The average of the two measurements was used in the final analysis. Paired sample t tests were used for statistical analyses.\n\n\nResults\n\nData from all 72 selected children were used in the final analysis. Sex and age distributions are shown in Table 2. The mean TSALD in the case and control groups was -6.22±2.49mm and 0.42±2.30mm, respectively. In the case group, canines, with the average eruption percentage of 65.82±13.00, were significantly behind the premolars which were on average 78.92±10.15 percent erupted (p<0.001). However, the average eruption percentage of canines (74.12±14.55) was very close to that of first premolars (75.47±11.60) in the control group (p=0.437) (Table 3). Therefore, in the case group, which had greater TSALD than controls, the first premolar teeth would erupt before the canines, while there was no priority in the control group. The difference in eruption percent of canines and first premolar teeth between cases and controls was statistically significant (p<0.001).\n\n\n\n\nDiscussion\n\nThis study was designed and conducted based on the hypothesis that the eruption sequence of the permanent canines and premolars of children with dental crowding differs from that of children without dental crowding. To the author’s knowledge, this is one of the first studies reporting a clear correlation between canine and premolar eruptive position in the mixed dentition stage and dental crowding. The results showed that there was a significant difference in the pre-eruptive positions of canine and first premolar teeth in cases with moderate to severe crowding compared to controls. The difference between the groups was large enough to be clinically detected. Clear clinical differences between canine and first premolar eruption order could be detected by a cursory assessment of panoramic radiographies and seemed to confirm the results. Therefore, canine and premolar eruption order assessed on radiographies can be used to identify children with high risk of developing moderate to severe crowding in mixed dentition.\n\nIn our study, we selected children from 8 to 10 years old age and compared the eruption percentage of two teeth, namely canine and first premolar in the mandible. The age group selected is corresponding to the mixed dentition age when most of orthodontic interceptions could be effective. Often crowding is more severe in the mandibular arch as there are several mechanisms to resolve the lack of space in the maxilla11–14.\n\nWe observed that patients with dental crowding presented a possible delay in the eruption of their permanent mandibular canines compared to their adjacent premolars. This observation was compatible with Bradley’s notation of retardation in the eruption time of mandibular canines when there was a lack of space13. Lange has also found that more crowding is observed with the eruption sequence of 4-3-5 compared with 3-4-5 (i.e. when the first premolar erupts before canine), a finding that is exactly the same as ours14.\n\nIt has been shown in the literature that the utilization of leeway space is indicative of crowding20,21. Moorrees also concludes that utilization of leeway space depends on the sequence of eruption and shedding of posterior teeth. We can therefore justify our findings on the basis that eruption of the first premolar before the canine teeth may result in using the leeway space inappropriately which increases the chance of development of dental crowding12.\n\nOur study suggests a practical approach for early identification of children susceptible to develop dental crowding. Just two teeth were assessed. That made this study different from other similar studies in which too many variables from several teeth were considered11,13. As a result, a simple significant difference was found in the present study that can act as a practical clue for clinicians.\n\nWe used TSALD measurements in the total arch as the indicator of crowding. This is another advantage of this study over the few similar ones that only calculated the crowding in the canine-premolar segment13. Most crowding in the mandible occurs in the incisor-canine segment and in central-lateral contact first and then migrates to posterior segments22,23.\n\nClinicians are often confronted with decisions concerning the choice of interceptive treatments for potential crowding during mixed dentitions. Extreme caution should be exercised in selecting patients that will truly benefit from interceptive procedures such as planned extractions. The findings of the present study can be used together with other clues to select suitable cases for such treatments.\n\n\nConclusion\n\nThe findings of this study showed that children whose first premolar teeth precede their canines in eruption are more likely to develop malocclusions related to TSALD later on. Therefore, routine screening of the panoramic views of children seeking orthodontic consultation in mixed dentition would be helpful in the diagnosis of children with the chance of developing moderate to severe crowding in their permanent dentition.\n\n\nData availability\n\nF1000Research: Dataset 1. Tooth size-arch length discrepancies and teeth eruption in children with and without dental crowding, 10.5256/f1000research.3196.d2772924\n\n\nConsent\n\nWritten informed consent for publication of clinical details was obtained from the parents of the children.", "appendix": "Author contributions\n\n\n\nThe study was designed by V.M. All authors participated in the data collection and analysis. The first draft was prepared by A.G. All authors read and approved the final draft.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was financially supported by the Orthodontics Research Center and the vice-chancellery for Research of Shiraz University of Medical Sciences.\n\n\nAcknowledgements\n\nThe proposal of this study was peer reviewed and approved by the Orthodontics Research Center’s main and ethical boards.\n\n\nReferences\n\nProffit WR, Fields HW, Moray LJ: Prevalence of malocclusion and orthodontic treatment need in the United States: estimates from NHANES III survey. Int J Adult Orthod Orthognath Surg. 1998; 13(2): 97–106. PubMed Abstract\n\nSeehra J, Newton JT, Dibiase AT: Interceptive orthodontic treatment in bullied adolescents and its impact on self-esteem and oral-health-related quality of life. Eur J Orthod. 2013; 35(5): 615–621. PubMed Abstract \| Publisher Full Text\n\nBresnahan BW, Kiyak HA, Masters SH, et al.: Quality of life and economic burdens of malocclusion in U.S. patients enrolled in Medicaid. J Am Dent Assoc. 2010; 141(10): 1202–12. PubMed Abstract \| Publisher Full Text\n\nKerosuo H, Väkiparta M, Nyström M, et al.: The seven-year outcome of an early orthodontic treatment strategy. J Dent Res. 2008; 87(6): 584–588. PubMed Abstract \| Publisher Full Text\n\nKing GJ, Spiekerman CF, Greenlee GM, et al.: Randomized clinical trial of interceptive and comprehensive orthodontics. J Dent Res. 2012; 91(7 Suppl): 59S–64S. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKing GJ, Hall CV, Milgrom P, et al.: Early orthodontic treatment as a means to increase access for children enrolled in Medicaid in Washington State. J Am Dent Assoc. 2006; 137(1): 86–94. PubMed Abstract \| Publisher Full Text\n\nBaccetti T, Franchi L, Kim LH: Effect of timing on the outcomes of 1-phase nonextraction therapy of Class II malocclusion. Am J Orthod Dentofacial Orthop. 2009; 136(4): 501–509. PubMed Abstract \| Publisher Full Text\n\nDale JG, Dale HC: Interceptive guidance of occlusion with emphasis on diagnosis. In: Graber TM, Vanaardsall RL, Vig K, editors. Orthodontics: current principles and techniques. 4th ed. Missouri: Elsevier Mosby. 2005; 405–489.\n\nShaweesh AI: Polymorphisms in sequence of permanent tooth emergence: a cross-sectional study on Jordanian children and adolescents. Acta Odontol Scand. 2013; 71(1): 32–7. PubMed Abstract \| Publisher Full Text\n\nAlmonaitiene R, Balciuniene I, Tutkuviene J: Standards for permanent teeth emergence time and sequence in Lithuanian children, residents of Vilnius city. Stomatologija 2012; 14(3): 93–100. PubMed Abstract\n\nSampson WJ, Richards LC: Prediction of mandibular incisor and canine crowding changes in the mixed dentition. Am J Orthod. 1985; 88(1): 47–63. PubMed Abstract \| Publisher Full Text\n\nMoorrees CF, Reed RB: Correlations among crown diameters of human teeth. Arch Oral Biol. 1964; 9(6): 685–697. PubMed Abstract \| Publisher Full Text\n\nBradley RE: The relationship between eruption, calcification and crowding of certain mandibular teeth. Angle Orthod. 1961; 31(4): 230–236. Reference Source\n\nLange GM: Correlation of sequence of eruption and crowding. Masters’ Thesis, Saint Louis University, St Louis, MO 63103, United States. 2011. Reference Source\n\nSharma R, Kumar S, Singla A: Prevalence of tooth size discrepancy among North Indian orthodontic patients. Contemp Clin Dent. 2011; 2(3): 170–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoyer RE: Handbook of orthodontics. Analysis of the dentition and occlusion. 4th ed. Year Book Medical Publishers; 1988. 235–8.\n\nTanaka MM, Johnston LE: The prediction of the size of unerupted canines and premolars in a contemporary orthodontic population. J Am Dent Assoc. 1974; 88(4): 798–801. PubMed Abstract \| Publisher Full Text\n\nAltherr ER, Koroluk LD, Phillips C: Influence of sex and ethnic tooth-size differences on mixed-dentition space analysis. Am J Orthod Dentofacial Orthop. 2007; 132(3): 332–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShumaker DB, Hadary EL: Roentgenographic study of eruption. J Am Dent Assoc. 1960; 61: 535–41.\n\nGianelly AA: Leeway space and the resolution of crowding in the mixed dentition. Semin Orthod. 1995; 1(3): 188–94. PubMed Abstract \| Publisher Full Text\n\nVyas MB, Hantodkar N: Resolving mandibular arch discrepancy through utilization of leeway space. Contemp Clin Dent. 2011; 2(2): 115– 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBernabé E, Villanueva KM, Flores-Mir C: Tooth width ratios in crowded and noncrowded dentitions. Angle Orthod. 2004; 74(6): 765–768. PubMed Abstract\n\nPuri N, Pradhan KL, Chandna A, et al.: Biometric study of tooth size in normal, crowded, and spaced permanent dentitions. Am J Orthod Dentofacial Orthop. 2007; 132(3): 279.e7–14. PubMed Abstract \| Publisher Full Text\n\nGolkari A, Moshkelgosha V, Khosravifard N: F1000Research: Dataset 1. Tooth size-arch length discrepancies and teeth eruption in children with and without dental crowding. Data Source" }	[ { "id": "5046", "date": "09 Jun 2014", "name": "Wael Sabbah", "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 very good paper, methodologically sound. The findings of the paper have the potential to allow early detection of malocclusion cases and would possibly benefit treatment planning for orthodontic cases. A minor change is needed. In the first paragraph the authors referred to 'NHANES III 1998'. The date for NHANES III is 1988-1994. If the authors are referring to another year or to NHANES in general they should amend the text and use the reference.", "responses": [] }, { "id": "5035", "date": "12 Jun 2014", "name": "Mahtab Nouri", "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 paper about the role of sequence and percentage of eruption of first premolar in comparison to canine on development of crowding in the mandible. This case control study has a sound research methodology. I wished that there was an explanation added about the dataset regarding the control group and the method of its TSALD calculation. The conclusion is justified.", "responses": [] } ]	1	https://f1000research.com/articles/3-122
https://f1000research.com/articles/3-76/v1	21 Mar 14	{ "type": "Case Report", "title": "A case of hypertrophic lupus erythematosus with negative CD123 staining and transepidermal elimination of elastin", "authors": [ "Matthew Hughes", "Jerad M. Gardner", "Ling Gao", "Matthew Hughes", "Jerad M. Gardner" ], "abstract": "We report the case of a 49-year-old male with clinical and histological findings consistent with hypertrophic lupus erythematosus (HLE). HLE must be clinically and histologically differentiated from keratoacanthoma, hypertrophic lichen planus, squamous cell carcinoma and plaque type psoriasis. CD123 positivity and transepidermal elimination of elastin have recently been reported as tools to distinguish HLE. Interestingly, in this case, biopsies of two separate lesions failed to reveal these two features. The etiology of this discrepancy is unknown and further studies are needed to clarify the utility of CD123 positivity and transepidermal elimination of elastin in the diagnosis of hypertrophic lupus erythematosus.", "keywords": [ "Hypertrophic lupus erythematosus (HLE) is a rare subset of discoid lupus erythematosus", "characterized by erythematous", "indurated", "verrucous papules and nodules located on sun-exposed areas. HLE must be clinically and histologically differentiated from keratoacanthoma", "hypertrophic lichen planus", "squamous cell carcinoma and plaque type psoriasis. CD123 positivity and transepidermal elimination of elastin have recently been reported to distinguish HLE1", "2." ], "content": "Introduction\n\nHypertrophic lupus erythematosus (HLE) is a rare subset of discoid lupus erythematosus, characterized by erythematous, indurated, verrucous papules and nodules located on sun-exposed areas. HLE must be clinically and histologically differentiated from keratoacanthoma, hypertrophic lichen planus, squamous cell carcinoma and plaque type psoriasis. CD123 positivity and transepidermal elimination of elastin have recently been reported to distinguish HLE1,2.\n\n\nReport of case\n\nA 49-year-old, unemployed, white male presented with a three-year history of an expanding “rash”. He reported no constitutional symptoms. He had previously been treated with oral prednisone and an unknown topical steroid without improvement and was off all medications at our initial visit. The patient had a past medical history of hepatitis C. He denied a family history of skin or autoimmune diseases. Laboratory work-up was significant for positive anti-nuclear antibodies and anti-Ro antibodies. Physical exam revealed multiple hyperkeratotic, verrucous papules and nodules with white, scaly, cribriform centers overlying patches of depigmentation, erythema and atrophy on his bilateral arms (Figure 1) and anterior legs. His face and scalp had several atrophic, depigmented patches. Two punch biopsies were obtained from separate lesions. Histological sections demonstrated an interface inflammatory pattern with deep peri-vascular and peri-appendageal lymphocytic infiltrate and rare plasma cells (Figure 2). A diagnosis of HLE was made. The patient was prescribed clobetasol ointment 0.05% twice daily. At the three month follow-up, there was improvement of the hypertrophic lesions. The patient was subsequently lost to follow-up.\n\nHypertrophic lupus erythematosus presenting as a verrucous plaque on the patient’s elbow\n\nHypertrophic lupus erythematosus displays epidermal acanthosis and expansion of follicular ostia with a superficial and deep perivascular and periappendageal intradermal lymphocytic infiltrate (hematoxylin and eosin, 40 × magnification).\n\n\nDiscussion\n\nHLE was first described by Bechet in 19403. Clinical diagnosis can be challenging as HLE can mimic psoriasis or even squamous cell carcinoma. Uitto et al. described two histological patterns of HLE One resembled hypertrophic lichen planus, while the other was similar to keratoacanthoma4. Daldon et al. found that transepidermal elimination of elastin was present in 14 cases of HLE1. Recently, Ko et al. reported that a band of CD123 positive cells at the dermal-epidermal junction was characteristic of five cases of HLE2.\n\nIn this patient, we examined these two recently described histologic features of HLE. Interestingly, both CD123 positivity and transepidermal elimination of elastin were not present in this case. However, the histological and clinical findings were most consistent with HLE. The etiology of this discrepancy is unknown and further studies are needed to clarify the utility of CD123 positivity and transepidermal elimination of elastin in the diagnosis of hypertrophic lupus erythematosus.\n\nThere is no definitive treatment for HLE. Options include topical or intralesional steroids, topical or oral retinoids, topical calcineurin inhibitors, thalidomide, hydroxychloroquine and surgical excision5,6. Winchester et al. reported on the efficacy TNF-alpha inhibitors7.\n\nThis case highlights the discrepancies of CD 123 positivity and transepidermal elimination of elastin in HLE.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nHughes – data collection, manuscript preparation\n\nGardner – data collection, manuscript preparation\n\nGao – manuscript preparation, oversight/supervision\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\nDaldon PE, Macedo de Souza E, Cintra ML: Hypertrophic lupus erythematosus: a clinicopathological study of 14 cases. J Cutan Pathol. 2003; 30(7): 443–8. PubMed Abstract \| Publisher Full Text\n\nKo CJ, Srivastava B, Braverman I, et al.: Hypertrophic lupus erythematosus: the diagnostic utility of CD123 staining. J Cutan Pathol. 2011; 38(11): 889–92. PubMed Abstract \| Publisher Full Text\n\nBechet PE: Hypertrophic lupus erythematosus. Arch Dermato Syph. 1940; 42: 211.\n\nUitto J, Santa-Cruz DJ, Eisen AZ, et al.: Verrucous lesion in patients with discoid lupus erythematosus. Clinical, histopathological and immunofluorescence studies. Br J Dermatol. 1978; 98(5): 507–520. PubMed Abstract\n\nAl-Mutairi N, Rijhwani M, Nour-Eldin O: Hypertrophic lupus erythematosus treated successfully with acitretin as monotherapy. J Dermatol. 2005; 32(6): 482–6. PubMed Abstract\n\nGambini D, Carrera C, Passoni E, et al.: Thalidomide treatment for hypertrophic cutaneous lupus erythematosus. J Dermatolog Treat. 2004; 15(6): 365–371. PubMed Abstract \| Publisher Full Text\n\nWinchester D, Duffin KC, Hansen C: Response to ustekinumab in a patient with both severe psoriasis and hypertrophic cutaneous lupus. Lupus. 2012; 21(9): 1007–10. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4229", "date": "14 Apr 2014", "name": "Victoria P. Werth", "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 a case of hypertrophic lupus erythematosus that is described as unusual in pathologic presentation. The title needs to indicate absence of transepidermal elimination of elastin. It is currently unclear if transepidermal elmination of elastin was present. The order of treatment described for HLE is confusing. One would start with hydroxychloroquine, then add quinacrine to hydroxychloroquine, with topicals as adjunctive therapy. Oral retinoids, thalidomide, and immuosuppressives would be options. Given that frequently there are multiple lesions that may actually koebnerize in a surgical scar, one would not include surgical excision as an option. The report cited in favor of TNF-alpha inhibitor is on ustekinumab, which is not a TNF inhibitor. This needs revision. Information about the antibody used for CD123 staining, as well as whether frozen or fixed tissue was used, is important. Anti-CD123 staining is not as good on fixed tissue. Were there any positive controls stained simultaneously?", "responses": [ { "c_id": "844", "date": "28 May 2014", "name": "Jerad Gardner", "role": "Reader Comment", "response": "Thank you for your commentary. The title has been changed to more clearly reflect the absence of transepidermal elimination in this case. The treatments listed in the report were a review of treatment options from the literature. They were not ordered as a suggested line of therapy. Ustekinumab has been reported to improve the plaques of hypertrophic lupus erythematosus. It is an inhibitor of IL-12 and IL-23. This fact has been corrected. We performed CD 123 staining on paraffin embedded tissue which is the method Ko et al. employed in her report on the novel use of CD 123 staining in hypertrophic lupus erythematosus. The second version of this report will contain a figure displaying positive control staining and negative CD 123 staining of the biopsy from this case." } ] }, { "id": "4445", "date": "29 Apr 2014", "name": "Theresa T. Lu", "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 report describes a case of hypertrophic lupus erythematosus based on clinical and histopathologic criteria that is negative for CD123 and elastin elimination. Negative data is important. However, as the emphasis is on the lack of CD123 and the lack of transepidermal elastin elimination, it would be good to show the negative results. For the CD123 stain, it would be good to show a positive control to make sure that the antibody really worked.", "responses": [ { "c_id": "843", "date": "28 May 2014", "name": "Jerad Gardner", "role": "Reader Comment", "response": "Thank you for your commentary. The second version of this report will contain a histological photo of the positive control and negative CD 123 staining of the biopsy specimen." } ] } ]	1	https://f1000research.com/articles/3-76
https://f1000research.com/articles/3-121/v1	02 Jun 14	{ "type": "Research Article", "title": "Yeast Augmented Network Analysis (YANA): a new systems approach to identify therapeutic targets for human genetic diseases", "authors": [ "David J. Wiley", "Ilona Juan", "Hao Le", "Xiaodong Cai", "Lisa Baumbach", "Christine Beattie", "Gennaro D'Urso", "David J. Wiley", "Ilona Juan", "Hao Le", "Xiaodong Cai", "Lisa Baumbach", "Christine Beattie" ], "abstract": "Genetic interaction networks that underlie most human diseases are highly complex and poorly defined. Better-defined networks will allow identification of a greater number of therapeutic targets.Here we introduce our Yeast Augmented Network Analysis (YANA) approach and test it with the X-linked spinal muscular atrophy (SMA) disease gene UBA1. First, we express UBA1 and a mutant variant in fission yeast and use high-throughput methods to identify fission yeast genetic modifiers of UBA1. Second, we analyze available protein-protein interaction network databases in both fission yeast and human to construct UBA1 genetic networks. Third, from these networks we identified potential therapeutic targets for SMA. Finally, we validate one of these targets in a vertebrate (zebrafish) SMA model. This study demonstrates the power of combining synthetic and chemical genetics with a simple model system to identify human disease gene networks that can be exploited for treating human diseases.", "keywords": [ "Many disease-associated genes have been identified", "yet most genetic diseases remain untreatable. One path to treatment is to develop extensive genetic networks in which human disease genes function (or dysfunction) and then target therapies to the genes identified in those networks. Eventually", "an individual’s own genotype for proteins in the network may also be considered in the therapeutic options. Developing gene networks is already widely recognized as a powerful approach to identify new drug targets1–6. Gene networks are based on the principle that networks contain proteins that interact (physically and/or functionally) and that these interactions govern most", "if not all", "cellular functions. Importantly", "gene network interactions are often conserved in different organisms even though the output of the networks may differ7. Candidate networks can be generated from studies in model organisms and then extrapolated to human cells. One can expect that", "for disease gene networks", "some of the interacting genes might modulate the disease phenotype thus representing potential therapeutic targets." ], "content": "Introduction\n\nMany disease-associated genes have been identified, yet most genetic diseases remain untreatable. One path to treatment is to develop extensive genetic networks in which human disease genes function (or dysfunction) and then target therapies to the genes identified in those networks. Eventually, an individual’s own genotype for proteins in the network may also be considered in the therapeutic options. Developing gene networks is already widely recognized as a powerful approach to identify new drug targets1–6. Gene networks are based on the principle that networks contain proteins that interact (physically and/or functionally) and that these interactions govern most, if not all, cellular functions. Importantly, gene network interactions are often conserved in different organisms even though the output of the networks may differ7. Candidate networks can be generated from studies in model organisms and then extrapolated to human cells. One can expect that, for disease gene networks, some of the interacting genes might modulate the disease phenotype thus representing potential therapeutic targets.\n\nThe direct identification of disease networks from studies in human cells is, of course, challenging because of the large number of potential proteins and incomplete knowledge of how the activity of any one protein may affect the network output. Analysis of networks would benefit from studies in a tractable model organism amenable to high throughput methods. Given that network interactions may be conserved between humans and model organisms, even while the outputs may differ, we hypothesized that we can identify human disease networks in the fission yeast Schizosaccharomyces pombe, a simple, genetically tractable model organism, coupled to existing protein-protein interaction databases. We can then transfer that network knowledge to human cells with the goal of identifying therapeutic target genes.\n\nHow valid is the approach of using fission yeast networks to identify human disease gene networks? First, the proteins that control most core cellular functions are in fact evolutionarily conserved, underscoring the “deep homology” that exists between all living organisms8,9. For example, in S. pombe >65% of the genome is orthologous to human (http://orthomcl.org/orthomcl/ and http://www.pombase.org). Second, many of the cellular processes implicated in human disease, e.g. vesicular transport, protein folding, metabolism, and RNA processing, are also evolutionarily conserved and highly interconnected10. Importantly, the value of a simple model organism for discovering “druggable” genetic pathways has recently been demonstrated for the budding yeast Saccharomyces cerevisiae11, and thus extending the study of human disease genes to fission yeast seems promising.\n\nOur approach (Figure 1) involves the expression of human disease associated genes in S. pombe and then the analysis of their effect on yeast fitness (growth). We performed high-throughput synthetic genetic array (SGA) screens to identify the fission yeast genetic modifiers that alter this effect, as measured by a simple yeast growth assay. The genetic modifiers are then assembled into human disease gene networks or clusters (using protein-protein interaction datasets from both S. pombe and humans). Any modifiers or genes in the networks or clusters represent potential therapeutic targets. This unbiased high-throughput approach could be widely and rapidly applied to many different disease-associated genes at relatively low cost.\n\n(1) Construction of query strains by integrating wildtype and mutant variants of the UBA1 gene into pombe. (2) Crossing the query strains (mutant or wildtype) to a complete non-essential deletion collection. (3) Bioinformatic analysis to generate the top ‘gene hits’. (4) Validation of the top hits in yeast. (5) Confirmation of the top hits in zebrafish.\n\nTo demonstrate the power of YANA, we considered spinal muscular atrophy, a common neurodegenerative disease affecting approximately 1/6000 births worldwide and the number one genetic cause of infantile death in the United States12. Most cases (>90%) are caused by deletion of SMN1, a gene encoding the survival motor neuron (SMN), a protein involved in the assembly of spliceosomal small nuclear ribonuceoproteins (snRNPs)13. In contrast, X-linked spinal muscular atrophy is caused by mutations in UBA1, a gene encoding the ubiquitin activating (UBA) enzyme 1, an E1 ubiquitin ligase14,15. Consistent with a role of ubiquitination in spinal muscular atrophy and SMN biology, it was recently shown that ubiquitin-mediated proteolysis regulates SMN stability16,17. Therefore, we systematically investigated the genetic network for UBA1 to better understand the potential connections between UBA1, its modifiers and the spinal muscular atrophy phenotype. Using YANA we identified several potential therapeutic targets and validated one of these targets in a SMA vertebrate (zebrafish) model.\n\n\nMethods\n\nUBA1 was cloned into a pENTR/D-TOPO vector (Life Technologies, Cat # K2400-20) from cDNA (Origene, Cat. # SC320329) following the manufacturer’s protocols. The following primers were used: forward primer: 5′- CACCATGTCCAGCTCGCCGC-3′; reverse primer: 5′- TCAGCGGATGGTGTATCGGAC-3′. Genetic insertion was confirmed by sequencing (http://sylvester.org/shared-resources/oncogenomics). UBA1 (G1617T), mutUBA1, was created by site directed mutagenesis using a QuikChange Lightning Site-Directed Mutagenesis kit (Agilent Technologies, Cat. # 210518; the detailed protocol is available in the kit). The primers used were: forward primer: 5′-GCAGCTGTGCGCCAAATTAATCCACATATCCGG-3′; reverse primer: 5′-CCGGATATGTGGATTAATTTGGCGCACAGCTGC-3′. LR Gateway reactions (Gateway Cloning Protocols: http://www.lifetechnologies.com/us/en/home/life-science/cloning/gateway-cloning/protocols.html#lr) were then performed to flip the UBA1 genes into destination vectors, to create N-terminal HA-tagged UBA1 under the control of the nmt1 promoter (LR Clonase II from Life Technologies, Cat # 11791020). The newly generated expression vectors were then integrated into an h- leu1-32 ura4-D18 Ade6-M210 S. pombe strain (PN572) to create a wtUBA1 query strain (h- integrated pjk148-nmt13X-HA-UBA1-nmt1term. leu1-32 ura4-D18 Ade6-M210) and a mutUBA1 query strain (h- integrated pjk148-nmt13X-HA-UBA1(G1617T)-nmt1term. leu1-32 ura4-D18 Ade6-M210). All media, growth conditions, and genetic manipulations were as previously described18.\n\nStrains containing wtUBA1 and mutUBA1 were grown exponentially in PMG media (Sunrise Scientific Cat. #2060, keeping the OD595 below 0.4) or eight generations at 32°C and then induced for expression of the HA-tagged wtUBA1 and mutUBA1 by washing the cells three times with sterile water to remove thiamine. Cells were then grown exponentially for 16 hours and then lysed using a FastPrep 120 bead beater (MP Biomedical), followed by boiling in sample buffer (2x Laemmli Sample Buffer, Bio-Rad Cat. #161-0737) and then clarified by centrifugation. The expression of UBA1 was confirmed by Western Blot analysis using an anti-HA antibody at 1:2000 dilution (Covance, Cat #MMS-101P, AB_10063488) and standard procedures19. Growth curve analysis was completed in the presence and absence of thiamine (final concentration, 15 µM; Sigma, Cat. #T4625). Removal of thiamine was achieved by washing the cells three times with sterile water. Cells were incubated in the absence of thiamine for 22 hours under exponential growth, diluted and 120 µl added to the Tecan Infinity F200 plate readers (starting OD595 of 0.05) for growth curve analysis with an n=5 for each sample. For validation experiments, cul3 and gsk3 were not pre-induced, while pub1 was pre-induced for 22 hours. The deletions of cul3, pub1, and gsk3 were confirmed by PCR analysis of genomic DNA from each strain, as previously described20.\n\nQuery strains were grown in liquid media and then pinned to agar in a 384-format using a RoToR HDA (Singer Instruments). The query was then crossed to the S. pombe haploid deletion library (Bioneer, Version 3.0 equivalent) on SPAS media (details can be found at http://www-bcf.usc.edu/~forsburg/media.html) using a modified SGA procedure21. For germination, four replicates of each cross were pinned to a 1536 format on selective PMG media containing thiamine, adenine (225 mg/L, Sigma Cat. #A8751), leucine (225 mg/L, Sigma Cat. #L8912), and the antibiotic G418 (150 µM, Gold Bio Cat. #G-1418) (PAUT+G418). After three days, the plates were then pinned to both PAUT+G418 (non-inducing) and PAU+G418 (inducing) plates. The colonies growing on the PAUT+G418 plates were documented on a flatbed scanner for the next three consecutive days. The PAU+G418 plates were grown for two days, and then re-pinned to fresh PAU+G418 plates and documented over the next three consecutive days. Based on the growth characteristics of the wtUBA1 strain, the plates were then pinned to fresh PAU+G418 plates and documented for an additional three days.\n\nThe documented plates were analyzed for ‘Hits’ representing a growth defect (SL) or growth suppressor (SS) using ScreenMill software22. The software is used to quantify colony size for each individual cross and then to normalize the quantified plates with and without replicate exclusion for each quadruplicate of the query crossed to a specific deletion strain. Data was then compared (between non-induced versus induced) and ranked in Excel (P≤0.05) (Microsoft). The orthologs were then identified based on curated data from PomBase (www.pombase.org; build 2013-11-11-v1), OrthoMCL (http://orthomcl.org/orthomcl/; Version 5), InParanoid8 (http://inparanoid.sbc.su.se/cgi-bin/index.cgi; Version 8.0), and Homologene (http://www.ncbi.nlm.nih.gov/homologene; build 67).\n\nThe S. pombe and human ‘Hits’ were analyzed in String (www.string-db.org) to map protein-protein interactions limited to data from experiments at the highest confidence level (0.900), and named pombe primary and human primary. The S. pombe primary was then extended by adding first degree neighbors to the original pombe ‘Hits’, keeping all interactions based on data from experiments at the highest confidence level (0.900). From the extended pombe network interaction map, new ‘Hits’ were extracted that corresponded to at least two previous ‘orphans’ (‘Hits’ that were not previously mapped) that interact through a nearest neighbor. All ‘Hits’ were then analyzed through PubMed (http://www.ncbi.nlm.nih.gov/pubmed/) for relevance in terms of SMA, with special focus on the ‘Hits’ in the interaction maps.\n\nTransgenic Tg(mnx1:0.6hsp70:GFP)os26 embryos that express GFP in ventrally projecting motor axons, referred to as Tg(mnx1:GFP) embryos, were used for all zebrafish experiments. Embryos were staged according to Fritz et al.23 All fish were grown and maintained in the Ohio State University (OSU) zebrafish facility following established protocols and OSU animal welfare guidelines as stated in Dr. Beattie's animal protocol (On file at OSU: 2009A0141-R1). Tg(mnx1:GFP) embryos were injected with 4 ng smn MO at the one- to 2-cell stage to knock down Smn as previously described24. At 10 hours post-fertilization (hpf) injected embryos (in their chorions) were placed in Petri dishes in fish water (60 µg/ml Instant Ocean® sea salts) containing compounds in 0.75% dimethyl sulfoxide (DMSO) or 0.75% DMSO only and incubated at 28.5°C in incubator until 28 hpf.\n\nTg(mnx1:GFP) embryos at 28 hpf were anesthetized with tricaine (250 µg/ml, Sigma A-5040) and fixed overnight at 4°C in 4% formaldehyde/PBS. After removing embryos from fix, they were mounted on glass coverslips for observation under a Zeiss Axioplan microscope. Motor axons innervating the mid-trunk (myotomes 6–15) on both sides of the fish were scored as described25.\n\nThe control toppedb458 embryos were treated and fixed as described above for Tg(mnx1:GFP) embryos; the topped mutants, however, were processed for znp1 antibody (Hybridoma Bank Cat#znp-1, AB_531910) labeling as previously described26 to visualize motor axons for scoring as they did not have the Tg(mnx1:GFP) in the background.\n\nSamples of zebrafish embryos injected with 4 ng smn MO at the one- to two-cell stage and uninjected embryos were treated from 10 to 28 hpf with DMSO or with DMSO and compounds as described above were collected for western blot.\n\nSamples were generated by boiling 25 identically treated embryos in 75 µl blending buffer (63 mM Tris (pH 6.8), 5 mM EDTA, 10% SDS). 10 µl (equivalent to 3 embryos or 75 µg of protein) were added to 10 µl of sample buffer (100 mM Tris (pH 6.8), 0.2% bromophenol blue, 20% glycerol, 200 mM dithiothreitol) and run on a 10% polyacrylamide gel, blotted to nitrocellulose, probed with mouse anti-Smn (1/500; MANSMA12, a gift from Dr. G.E. Morris or anti HuD (1/1000; Santa Cruz Cat#sc-28299, AB_627765) and detected by chemiluminescence of bound HRP-conjugated mouse antibody. Blots were stripped and re-probed with mouse anti-β-actin (1/1000; Santa Cruz Cat #sc-47778, AB_626632).\n\n\nResults\n\nWe expressed the human wildtype UBA1 (wtUBA1) and the disease-causing variant 1617UBA1 (mutUBA1) in S. pombe under the control of a regulatable nmt1 promoter27. We identified strains with stable integrated human UBA1 genes expressing equivalent levels of UBA1 proteins (Figure 2A). Under non-inducing conditions (+ thiamine) all UBA1-containing strains had a similar growth rate to the control non-UBA1-containing wildtype strain (Figure 2B). We then found that cells expressing human wtUBA1 (fission yeast cells grown in media lacking thiamine to induce UBA1 expression), experienced a decrease in cell growth. No growth defect was observed for yeast cells expressing mutUBA1 when compared to the control (Figure 2B).\n\nWildtype and mutant forms of UBA1 were integrated into fission yeast under the control of the nmt1 thiamine-repressible promoter. (A) Western blot analysis of protein expression levels shows that both forms of UBA1 (wildtype and mutant) are expressed to equivalent levels. (B) Expression of wild-type UBA1 inhibits S. pombe cell growth.\n\nWe set out to identify yeast gene mutations that either enhance or suppress cell growth dependent upon expression of either wt or mutUBA1 (i.e., epistatic modifiers). We used automated genetics to introduce the wt and mutUBA1 genes into a deletion strain collection corresponding to more than 90% of the non-essential genes in S. pombe (ca. 80% of the complete genome). We then tested each of the >7000 unique strains expressing either wt or mutUBA1 combined with a specific gene deletion, for growth properties. Each strain was scored for significance of growth difference compared to a control, and the top hits (P<0.05) were assembled into a database (Table 1). The categories of growth phenotypes were those that were synthetic lethal for wildtype or mutant (wt-SL, or mut-SL) or synthetically suppressed for wildtype or mutant (wt-SS, or mut-SS). We identified 173 UBA1 or mutUBA1 modifiers. Notably, 145 of the modifiers (83.8%) were orthologous to human genes, and thus are potential drug targets to modify the SMA phenotype.\n\nThe tops hits for wtUBA1 SL, wtUBA1 SS, mutUBA1 SL, and mutUBA1 SS were identified by bioinformatic analysis as described in the materials and methods section.\n\nTo generate our candidate human disease gene networks (that we shall call network clusters) we relied on two complementary approaches. Both approaches rely on networks constructed from published protein-protein interaction data from either human (Figure 3 and Figure S3) or fission yeast (Figure 4, Figure S1 and Figure S2). By including two types of genetic interactions, synthetic lethals or synthetic suppressors, yeast augmented network analysis (YANA), yields the most complete unbiased list of potential modifiers. In a first approach, we converted all 145 S. pombe modifier genes to their corresponding human orthologs. All the data were combined to create network cluster diagrams. Individual datasets were delineated using a color key. (Note that we designate each interaction as a synthetic lethal or synthetic suppressor a priori, and include them both as it is difficult to predict which of these classes will have potential therapeutic value). Using the highest confidence (0.9) experimentally confirmed protein-protein interaction data (from www.string-db.org), we used cytoscape to draw network cluster modules that include all modifier genes. We identified 22 clusters using this approach (Figure S3). We then limited the number of networks by focusing on those modifiers that shared at least two interactions with other modifiers shrinking the number of clusters to 10 (Figure 3). Within our human primary network clusters, we identified several genes that represent compelling therapeutic targets. These included GSK3 (Glycogen Synthase Kinase-3), CUL3 (Cullin-3) and NEDD4 (E3 ubiquitin ligase) along with SUMO1 (Small Ubiquitin Modifier-1) and HDAC (Histon deacetylase), a known chromatin modifier. Interestingly, GSK3 inhibitors increase SMN levels in spinal muscular atrophy patient-derived fibroblasts and mouse motor neurons28. The HDAC inhibitor Trichostatin A increases SMN expression and survival in a mouse model of spinal muscular atrophy29. Therefore, evidence already exists that genes within our clusters affect SMA biology.\n\nHuman orthologs of yeast genes identified from four independent SGA screens were used to map the corresponding protein-protein interaction network. Only proteins that have experimentally verified data (STRING) supporting a physical interaction with at least two additional genes/proteins identified in the UBA1-SGA screens are displayed. Of particular interest is the discovery of NEDD4, CUL3, GSK3, SUMO and HDAC in distinct network clusters. Proteins of interest are highlighted by a larger node size and indicated with an arrow.\n\nPrimary yeast gene hits from all four SGA screens were combined and mapped based on highest confidence experimentally confirmed protein-protein interaction data. For this extended analysis, we also included first-degree neighbors of modifiers that were not directly identified in the SGA screens. Only selected gene clusters are shown here; the complete extended network is shown in Figure S1. When extrapolated to human, the yeast extended network analysis identified Csn5 and Cul3, two modifiers also identified in our human UBA1 primary gene interaction network (boxed; see also Figure 3).\n\nIn our second approach, we constructed networks from just the S. pombe modifier genes using high confidence S. pombe protein-protein interaction data from STRING. For this analysis, we identified network clusters that were expanded to include all the so-called “first-degree (1°) interacting proteins”, proteins not identified in our SGA screens but that physically interact with a genetically identified modifier (Figure S1). The rationale underlying our approach to generate larger clusters comes from the facts that the 1° neighbor may be an essential gene (and therefore it is not present in our deletion library) or may fail to be detected in our screens for other, unknown reasons. This “Cluster with 1° neighbor” analysis reveals new interactions not observed in the yeast primary UBA1 gene interaction network (Figure S2). A few examples of these types of clusters are shown in Figure 4. The clusters include several genes all of which were uniquely identified in the wildtype screen, including both synthetic lethals and synthetic suppressors. Of particular interest is the cluster of proteins surrounding Skp1(S-phase kinase associated protein-1), a protein that interacts and stablilizes F-box proteins, additional F-box proteins, and proteins involved in cullin deneddylation and neddylation. In addition, we found that the RNA-binding protein Cwf2 interacts directly with three of our modifiers, and the plastin ortholog Fim1, interacts with two. Interestingly, Plastin 3 (PLS3) has been identified as a disease modifier in animal models of spinal muscular atrophy30–32.\n\nWe then identified correlations between the extended S. pombe network clusters (Figure 4) and the human primary network clusters (Figure 3). In both networks, we identified components of the COPS complex, a known regulator of E3 cullins, and the E3 cullin Cul3. Note that Cul3 was not present in our primary S. pombe network (Figure S2) yet was present in the more extended network clusters that incorporate 1° neighbors; this illustrates the importance of adding 1° neighbors to our analysis of the S. pombe data.\n\nTogether, these data suggest a prominent role for E3 ubiquitin ligases in modulating UBA1 and potentially SMN1, making Cul3 the primary target for further analysis. The E3 ubiquitin ligase might inhibit SMN function, as ubiquitination often leads to protein degradation. If so, we reasoned that inhibition of E3 ubiquitin ligase might enhance SMN activity and thus suppress loss of function mutations in SMN1. To directly test the hypothesis that inhibition of E3 ubiquitin ligases can suppress a SMN1-spinal muscular atrophy phenotype, we turned to a vertebrate model of SMN deficiency previously reported in zebrafish33. Knockdown of Smn in zebrafish embryos causes developmental defects in motor neuron axonal outgrowth that include truncations and abnormal branching of neurons24. Motor neuron axon defects can be corrected by injection of mRNAs encoding wildtype human SMN25. We first confirmed that injection of smn morpholino (MO) into wildtype zebrafish causes severe motor axon abnormalities as compared to control uninjected embryos (Figure 5A and 5B). We then tested whether an inhibitor of E3 ubiquitin ligases (MLN4294) would suppress the motor axon abnormalities. Addition of the Nedd8-E1 activating enzyme inhibitor MLN4294 (that blocks cullin-RING E3 ligases), at concentrations ranging from 10 µM to 15 µM, caused a concentration-dependent reduction in the degree of abnormal motor axon branching (Figure 5A and 5B). At 15 µM inhibitor we found motor axon abnormalities that were completely suppressed: defects were not significantly different to those observed in uninjected embryos (P=0.379). At higher concentrations of drug (20 µM), the rescue was less pronounced (Table 2) probably because higher levels of drug led to defects in development. Thus, E3 ligase inhibition can rescue neuronal defects caused by Smn protein depletion in zebrafish. In control experiments, MLN4294 failed to rescue the zebrafish mutant topped (Figure 6), a mutant defective in neuronal axon guidance34. We also tested a drug that inhibits sumolyation corresponding to an unrelated target (SUMO) identified in our human primary network clusters (Figure 3); addition of this compound had no effect on the spinal muscular atrophy model (Figure S6) necessarily surprising since pmt3, which encodes the yeast ortholog of SUMO1, has the opposite effect of cul3 when deleted. Having been identified as a synthetic suppressor rather than a synthetic lethal, perhaps compounds that activate rather than inhibit sumolyation would be beneficial as a therapeutic.\n\n(A) Representative lateral views of motor axons in Tg(mnx1:GFP) zebrafish embryos expressing GFP in motor neurons and injected with control MO and then grown in 10 or 15 µM MLN4294. (B) Quantification of the effects of MLN4294 on motor axon development in zebrafish. Motor axons were scored in Tg(mnx1:GFP) embryos injected with control MO, and subsequently (10 hrs post-injection) incubated in 10 or 15 µM MLN4294. Embryos were classified as severe, moderate, mild, or no defects based on the severity of motor axon defects, and the percentage of each group is shown. Data in all graphs are represented as mean and SEM. (C) Western blot analysis of Smn and HuD protein following treatment with MLN4294. Quantification of the results is shown in Figure S5.\n\n(A) Quantification of the effects of MLN4294 on motor axon guidance in the zebrafish topped mutant. Motor axons were scored in embryos injected with DMSO, 10 or 15 µM MLN4294. Embryos were classified as severe, moderate, mild, or no defects based on the severity of motor axon defects, and the percentage of each group is shown. Data in all graphs are represented as mean and SEM. (B) Experimental data.\n\nTo determine how E3 ubiquitin ligases might influence Smn1 function, we examined the protein levels for both Smn and HuD. HuD is a known SMN interacting protein that binds RNAs controlling their translation and stability and functions in neural development and plasticity34–36. We observed greater than a 2-fold increase in both Smn and Hud protein levels following treatment with MLN4294, a NEDD8 activating enzyme (NAE) inhibitor that prevents activation of E3 ligases (Figure 5C and Figure S5). It is therefore possible that rescue of the neuronal defects in our zebrafish model might involve stabilization of Smn as well as of additional proteins within the Smn network.\n\n\n\n\nDiscussion\n\nIn this paper we introduce YANA, a fission yeast genetic assay that that when applied to a specific human disease gene can leverage protein-protein interaction data to characterize human disease networks. From these networks, we can identify and prioritize genetic pathways likely to modify the disease-associated gene activity, and predict those genes that can be exploited as therapeutic targets. Using UBA1, a gene associated with X-linked spinal muscular atrophy (XL-SMA), we found a high degree of homology between our yeast modifiers and their corresponding human genes (>80%), underscoring the extensibility of the assay. Specifically, we identified several network clusters and high priority targets for therapeutic intervention, including GSK3 and HDAC both of which are potential therapeutic targets for SMA. In addition, we found two novel related targets Cul3 (a cullin required for E3 ubiquitin ligase activity) and NEDD4, an E3 ubiquitin ligase.\n\nOur yeast system, with its small but complex eukaryotic genome, and complete deletion library, is unique in allowing unbiased genome-wide screening of deletions that alter human disease gene activity. Moreover, YANA can be applied to any human gene, regardless of the phenotype or availability of endogenous mutations. The number of candidate genes identified by YANA for UBA1 represents a fraction of the approximately 30,000 genes in the human genome, providing a significant enrichment of potential modifiers. Therefore, YANA offers a simple, cost-effective, and relatively rapid technology that could be applied to all human genetic diseases.\n\n\nData availability\n\nF1000Research: Dataset 1. Growth rate data of expression of UBA1 in S. pombe, 10.5256/f1000research.4188.d2850537.", "appendix": "Author contributions\n\n\n\nG.D. conceived the project. D.W., I.J., L.T.H performed all experiments. X.C. assisted in the bioinformatics analysis. L.B. identified the UBA1 mutation linked to XL-SMA and was co-PI on the MDA grant that supported this research. C.B. was responsible for designing the zebrafish experiments and scoring the results.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a grant from the Muscular Dystrophy Association (MDA186435). (Christine Beattie grant support RO1NS050414 (C.E.B.) with additional support from P30NS045758).\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 Christian Romero for his expert technical assistance. Special thanks to W. Burhans, C. Boone, B. Andrews, M. Constanzo, C. Myers, C. Nislow, L. Hartwell, J. Dallman, and P. Goldschmidt for their encouragement and discussions during the early phase of these studies.\n\n\nSupplementary materials\n\nThe full protein-protein interaction network including first-degree neighbors for all fission yeast gene modifiers identified from our two independent SGA screens.\n\nFission yeast epistatic modifiers from all four SGA datasets were combined and used to map a protein-protein interaction network using the highest confidence protein interaction data (STRING-db).\n\nHuman orthologs of yeast genes identified from four independent SGA screens were used to map the corresponding protein-protein interaction network.\n\nGrowth curves of the 3 double mutants confirmed our original SGA data showing that the deletions of cul3 and gsk3 are synthetically lethal with OE of UBA1, and the deletion of pub1 (NEDD4) suppresses the toxicity associated with OE of wildtype UBA1. For gsk3 and cul3, the cells were washed to remove thiamine and immediately placed in the plate reader whilst for pub1, cells were first pre-induced for 22 hrs. Deletions of cul3, pub1, and gsk3 were confirmed by PCR.\n\nQuantification of western blot results shown in Figure 5C. Treatment of zebrafish embryos with MLN4294 results in approximately a 3-fold increase in both Smn and HuD levels 18 hrs following treatment.\n\nA) MLN4924 treatment in smn MO zebrafish. B) MLN4924 treatment in topped mutant zebrafish. C) Anacardic acid treatment in smn MO zebrafish.\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\nMaxwell CA, Benítez J, Gómez-Baldó L, et al.: Interplay between BRCA1 and RHAMM regulates epithelial apicobasal polarization and may influence risk of breast cancer. PLoS Biol. 2011; 9(11): e1001199. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPujana MA, Han JD, Starita LM, et al.: Network modeling links breast cancer susceptibility and centrosome dysfunction. Nat Genet. 2007; 39(11): 1338–1349. PubMed Abstract \| Publisher Full Text\n\nMasoudi-Nejad A, Mousavian Z, Bozorgmehr JH: Drug-Target and disease networks: polypharmacology in the post-genomic era. In Silico Pharmacology. 2013; 1: 17. Publisher Full Text\n\nZheng H, Fridkin M, Youdim M: From single target to multitarget/network therapeutics in Alzheimer’s therapy. Pharmaceuticals (Basel). 2014; 7(2): 113–135. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVitali F, Mulas F, Marini P, et al.: Network-based target ranking for polypharmacological therapies. AMIA Jt Summits Transl Sci Proc. 2013; 2013: 168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcGary KL, Park TJ, Woods JO, et al.: Systematic discovery of nonobvious human disease models through orthologous phenotypes. Proc Natl Acad Sci U S A. 2010; 107(14): 6544–6549. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShubin N, Tabin C, Carroll S: Deep homology and the origins of evolutionary novelty. Nature. 2009; 457(7231): 818–823. PubMed Abstract \| Publisher Full Text\n\nShubin N, Tabin C, Carroll S: Fossils, genes and the evolution of animal limbs. Nature. 1997; 388(6643): 639–648. PubMed Abstract \| Publisher Full Text\n\nCostanzo M, Baryshnikova A, Bellay J, et al.: The genetic landscape of a cell. Science. 2010; 327(5964): 425–431. PubMed Abstract \| Publisher Full Text\n\nTardiff DF, Jui NT, Khurana V, et al.: Yeast reveal a “druggable” Rsp5/Nedd4 network that ameliorates alpha-synuclein toxicity in neurons. Science. 2013; 342(6161): 979–983. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZanetta C, Riboldi G, Nizzardo M, et al.: Molecular, genetic and stem cell-mediated therapeutic strategies for spinal muscular atrophy (SMA). J Cell Mol Med. 2014; 18(2): 187–196. PubMed Abstract \| Publisher Full Text\n\nBattle DJ, Kasim M, Yong J, et al.: The SMN complex: an assembly machine for RNPs. Cold Spring Harb Symp Quant Biol. 2006; 71: 313–320. PubMed Abstract \| Publisher Full Text\n\nDlamini N, Josifova DJ, Paine SM, et al.: Clinical and neuropathological features of X-linked spinal muscular atrophy (SMAX2) associated with a novel mutation in the UBA1 gene. Neuromuscul Disord. 2013; 23(5): 391–398. PubMed Abstract \| Publisher Full Text\n\nRamser J, Ahearn ME, Lenski C, et al.: Rare missense and synonymous variants in UBE1 are associated with X-linked infantile spinal muscular atrophy. Am J Hum Genet. 2008; 82(1): 188–193. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChang HC, Hung WC, Chuang YJ, et al.: Degradation of survival motor neuron (SMN) protein is mediated via the ubiquitin/proteasome pathway. Neurochem Int. 2004; 45(7): 1107–1112. PubMed Abstract \| Publisher Full Text\n\nBurnett BG, Muñoz E, Tandon A, et al.: Regulation of SMN protein stability. Mol Cell Biol. 2009; 29(5): 1107–1115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoreno S, Klar A, Nurse P: Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Method Enzymol. 1991; 194: 795–723. PubMed Abstract \| Publisher Full Text\n\nLaemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970; 227(5259): 680–5. PubMed Abstract \| Publisher Full Text\n\nKim DU, Hayles J, Kim D, et al.: Analysis of a genome-wide set of gene deletions in the fission yeast Schizosaccharomyces pombe. Nat Biotechnol. 2010; 28(6): 617–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDixon SJ, Fedyshyn Y, Koh JL, et al.: Significant conservation of synthetic lethal genetic interaction networks between distantly related eukaryotes. Proc Natl Acad Sci U S A. 2008; 105(43): 16653–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDittmar JC, Reid RJ, Rothstein R: ScreenMill: a freely available software suite for growth measurement, analysis and visualization of high-throughput screen data. BMC Bioinformatics. 2010; 11: 353. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFritz A, Rozowski M, Walker C, et al.: Identification of selected gamma-ray induced deficiencies in zebrafish using multiplex polymerase chain reaction. Genetics. 1996; 144(4): 1735–45. PubMed Abstract \| Free Full Text\n\nMcWhorter ML, Monani UR, Burghes AH, et al.: Knockdown of the survival motor neuron (Smn) protein in zebrafish causes defects in motor axon outgrowth and pathfinding. J Cell Biol. 2003; 162(5): 919–931. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarrel TL, McWhorter ML, Workman E, et al.: Survival motor neuron function in motor axons is independent of functions required for small nuclear ribonucleoprotein biogenesis. J Neurosci. 2006; 26(43): 11014–11022. PubMed Abstract \| Publisher Full Text\n\nHao le T, Duy PQ, Jontes JD, et al.: Temporal requirement for SMN in motoneuron development. Hum Mol Genet. 2013; 22(13): 2612–2625. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaundrell K: nmt1 of fission yeast. A highly transcribed gene completely repressed by thiamine. J Biol Chem. 1990; 265(19): 10857–64. PubMed Abstract\n\nMakhortova NR, Hayhurst M, Cerqueira A, et al.: A screen for regulators of survival of motor neuron protein levels. Nat Chem Biol. 2011; 7(8): 544–552. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAvila AM, Burnett BG, Taye AA, et al.: Trichostatin A increases SMN expression and survival in a mouse model of spinal muscular atrophy. J Clin Invest. 2007; 117(3): 659–671. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAckermann B, Kröber S, Torres-Benito L, et al.: Plastin 3 ameliorates spinal muscular atrophy via delayed axon pruning and improves neuromuscular junction functionality. Hum Mol Genet. 2013; 22(7): 1328–1347. PubMed Abstract \| Publisher Full Text\n\nOprea GE, Kröber S, McWhorter ML, et al.: Plastin 3 is a protective modifier of autosomal recessive spinal muscular atrophy. Science. 2008; 320(5875): 524–527. PubMed Abstract \| Publisher Full Text\n\nHao le T, Wolman M, Granato M, et al.: Survival motor neuron affects plastin 3 protein levels leading to motor defects. J Neurosci. 2012; 32(15): 5074–5084. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeattie CE, Carrel TL, McWhorter ML: Fishing for a mechanism: using zebrafish to understand spinal muscular atrophy. J Child Neurol. 2007; 22(8): 995–1003. PubMed Abstract \| Publisher Full Text\n\nRodino-Klapac LR, Beattie CE: Zebrafish topped is required for ventral motor axon guidance. Dev Biol. 2004; 273(2): 308–320. PubMed Abstract \| Publisher Full Text\n\nFallini C, Bassell GJ, Rossoll W: Spinal muscular atrophy: the role of SMN in axonal mRNA regulation. Brain Res. 2012; 1462: 81–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHubers L, Valderrama-Carvajal H, Laframboise J, et al.: HuD interacts with survival motor neuron protein and can rescue spinal muscular atrophy-like neuronal defects. Hum Mol Genet. 2011; 20(3): 553–579. PubMed Abstract \| Publisher Full Text\n\nWiley DJ, Ilona J, Hao LT, et al.: Growth rate data of expression of UBA1 in S. pombe. F1000Research. Data Source" }	[ { "id": "4970", "date": "05 Jun 2014", "name": "Corey Nislow", "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 addresses the vexing problem of how to unravel the complexity of human diseases. Such diseases often involve interactions between genes, and the question is therefore well-suited to an automated genetic approach, such as Synthetic Genetic Array technology, which was pioneered in bakers yeast and more recently adopted to another well-characterized model system, namely S. pombe. In this study the authors examine conserved genetic interactions between a fission yeast genes and a human query gene (both wildtype and loss-of-function mutants) then combine their experimental data with bioinformatics to characterize the network of genetic interactions and protein-protein interactions. They then leverage this network to identify interacting partners that comprise novel therapeutic targets. The example presented in this paper is the human UBA1 gene which encodes a gene that, when mutated, results in human X-linked spinal muscular atrophy disease. They perform whole genome genetic interaction mapping for digenetic interactions with both wild type and mutant versions of the gene, build a network with the observed interactions, and then expand that primary network by adding the known partners that interact with members of the primary network. They then turned to the zebrafish vertebrate model system to validate their newly identified therapeutic targets. This study adds another tool to the armamentarium of systems biology approaches with which the complexity of human diseases can be dissected. Because the system is new in the sense that fission yeast has been largely unexplored (compared to its budding yeast relative from which it has diverged by several hundred million years) the observations on fission yeast, when combined with those from other model systems, provide an attractive complementary approach to understanding complex human biology. Given that this study builds on earlier approaches, the authors should cite the previously published studies on genetic interactions, chemical genetic interactions and protein-protein interactions using budding yeast, fly and other models. Similarly, the comprehensive gene expression profiling approaches used in cultured mammalian cells as well as more recent whole genome RNA interference and CRISPR-based disruption screens should be cited. From a methodological perspective, in the interaction maps presented in Figures 3 and 4 it would be useful to describe or cite the methods used to define the links between genes. In the Results, where the authors describe the results of the different screens which include wild type and mutant human protein interactions as well as primary and secondary interactions it would be instructive to provide a graphic or flow chart that depicts how all the key pieces of data from each independent screen in fission yeast (and their analysis) were collapsed and presented together. It would also be instructive to provide background information or appropriate citations on the two compounds used in the zebrafish experiments, specifically their known or suspected mechanisms of action as well as the data to support these mechanisms. In the Discussion, the authors have room to put the YANA screening platform into context, comparing it to other orthologous genome-wide approaches. A key innovation in this study arises from the unique approach used in analyzing the data. Specifically, the conserved yeast genes that modify the phenotype induced by expressing a human gene of interest are assembled into human disease gene networks which are then augmented by additional protein-protein interaction data. Augmenting traditional network analysis in this manner may permit the identification of new therapeutic targets and target pathways. In summary, this comprehensive, genome-wide work represents an important addition to the systems biology toolkit to understand the complexities of human biology and the impact of this and subsequent data sets will surely increase as additional screens are added to the compendium. Minor points Western blot and growth curve: \"or eight generations\" is a typo, also the type of media the cells were grown in “exponentially for 16 hours” should be listed. In Figure 2 it is not clear which band the tubulin arrow is pointing to- the loading control bands should be indicated directly on the blot/gel. Also, it is apparent from this loading control that the amount of UPA1 expressed is extremely high, the authors should comment on the fact that the overexpressing lanes are actually UNDERloaded relative to the non-expressing lanes. Third paragraph of the Results: the software programme cytoscape should include a reference. Third paragraph of the Results: histone is misspelled \"histon\". Sixth paragraph of the Results: the authors should clearly distinguish between the published data in zebrafish and their new data collected in this study Sixth paragraph of the Results: as mentioned be inhibitor’s mechanism of action should be described in more detail Second paragraph of the discussion: approximately 30,000 genes in the human genome should be cited and probably corrected to the conventional number of 21,000.", "responses": [] }, { "id": "5211", "date": "23 Jun 2014", "name": "Jasper Rine", "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 Wiley et al manuscript describes a beautiful synthesis of contemporary genetic approaches to, with astonishing efficiency, identify lead compounds for therapeutic approaches to a serious human disease. I believe the importance of this paper stems from the applicability of the approach to the several thousand of rare human disease genes that Next-Gen sequencing will uncover in the next few years and the challenge we will have in figuring out the function of these genes and their resulting defects. This work presents a paradigm that can be broadly and usefully applied. In detail, the authors begin with gene responsible for X-linked spinal muscular atrophy and express both the wild-type version of that human gene as well as a mutant form of that gene in S. pombe. The conceptual leap here is that progress in genetics is driven by phenotype, and this approach involving a yeast with no spine or muscles to atrophy is nevertheless and N-dimensional detector of phenotype. The study is not without a small measure of luck in that expression of the wild-type UBA1 gene caused a slow growth phenotype which the mutant did not. Hence there was something in S. pombe that could feel the impact of this protein. Given this phenotype, the authors then went to work and using the power of the synthetic genetic array approach pioneered by Boone and colleagues made a systematic set of double mutants combining the human expressed UBA1 gene with knockout alleles of a plurality of S. pombe genes. They found well over a hundred mutations that either enhanced or suppressed the growth defect of the cells expressing UBI1. Most of these have human orthologs. My hunch is that many human genes expressed in yeast will have some comparably exploitable phenotype, and time will tell. Building on the interaction networks of S. pombe genes already established, augmenting these networks by the protein interaction networks from yeast and from human proteome studies involving these genes, and from the structure of the emerging networks, the authors deduced that an E3 ligase modulated UBA1 and made the leap that it therefore might also impact X-linked Spinal Muscular Atrophy. Here, the awesome power of the model organism community comes into the picture as there is a zebrafish model of spinal muscular atrophy. The principle of phenologs articulated by the Marcotte group inspire the recognition of the transitive logic of how phenotypes in one organism relate to phenotypes in another.\n\nWith this zebrafish model, they were able to confirm that an inhibitor of E3 ligases and of the Nedd8-E1 activating suppressed the motor axon anomalies, as predicted by the effect of mutations in S. pombe on the phenotypes of the UBA1 overexpression. I believe this is an important paper to teach in intro graduate courses as it illustrates beautifully how important it is to know about and embrace the many new sources of systematic genetic information and apply them broadly.", "responses": [] } ]	1	https://f1000research.com/articles/3-121
https://f1000research.com/articles/3-120/v1	29 May 14	{ "type": "Research Article", "title": "Factors that contribute to social media influence within an Internal Medicine Twitter learning community", "authors": [ "Tejas Desai", "Manish Patwardhan", "Hunter Coore", "Manish Patwardhan", "Hunter Coore" ], "abstract": "Medical societies, faculty, and trainees use Twitter to learn from and educate other social media users. These social media communities bring together individuals with various levels of experience. It is not known if experienced individuals are also the most influential members. We hypothesize that participants with the greatest experience would be the most influential members of a Twitter community.We analyzed the 2013 Association of Program Directors in Internal Medicine Twitter community. We measured the number of tweets authored by each participant and the number of amplified tweets (re-tweets). We developed a multivariate linear regression model to identify any relationship to social media influence, measured by the PageRank.Faculty (from academic institutions) comprised 19% of the 132 participants in the learning community (p < 0.0001). Faculty authored 49% of all 867 tweets (p < 0.0001). Their tweets were the most likely to be amplified (52%, p < 0.01). Faculty had the greatest influence amongst all participants (mean 1.99, p < 0.0001). Being a faculty member had no predictive effect on influence (β = 0.068, p = 0.6). The only factors that predicted influence (higher PageRank) were the number of tweets authored (p < 0.0001) and number of tweets amplified (p < 0.0001)The status of “faculty member” did not confer a greater influence. Any participant who was able to author the greatest number of tweets or have more of his/her tweets amplified could wield a greater influence on the participants, regardless of his/her authority.", "keywords": [ "A number of medical societies", "faculty members", "and physician-trainees use social media", "specifically Twitter", "to learn from and educate other social media users1", "2. These social media communities offer a new and exciting medium by which knowledge can be shared and transmitted3. These communities bring together individuals/organizations with various levels of experience4", "5. In the traditional learning model", "the learners/students are aware of the authority of the teacher. In Twitter learning communities", "however", "there are many teachers whose levels of experience can vary. As a result", "individuals who participate in Twitter learning communities will be learning from multiple teachers of different levels of experience. This variety can pose a problem because inexperienced individuals can exert a great influence over learners. Although experienced teachers are increasingly participating in Twitter learning communities", "whether they are also the most influential members within the learning community is unknown3", "4", "6. We hypothesize that participants with the greatest experience would be the most influential members of one such Twitter learning community." ], "content": "Introduction\n\nA number of medical societies, faculty members, and physician-trainees use social media, specifically Twitter, to learn from and educate other social media users1,2. These social media communities offer a new and exciting medium by which knowledge can be shared and transmitted3. These communities bring together individuals/organizations with various levels of experience4,5. In the traditional learning model, the learners/students are aware of the authority of the teacher. In Twitter learning communities, however, there are many teachers whose levels of experience can vary. As a result, individuals who participate in Twitter learning communities will be learning from multiple teachers of different levels of experience. This variety can pose a problem because inexperienced individuals can exert a great influence over learners. Although experienced teachers are increasingly participating in Twitter learning communities, whether they are also the most influential members within the learning community is unknown3,4,6. We hypothesize that participants with the greatest experience would be the most influential members of one such Twitter learning community.\n\n\nMaterials and methods\n\nWe analyzed Twitter messages (tweets) from the 2013 Association of Program Directors in Internal Medicine meeting. This meeting was held from 28 April to 1 May 2013 and brought together residents and chief residents in Internal Medicine with faculty members and program directors. The Alliance for Academic Internal Medicine (AAIM) organized the meeting (http://www.im.org). The AAIM is a consortium of five academically focused organizations that represent Internal Medicine in the United States: 1) the Association of Professors in Medicine, 2) Association of Program Directors in Internal Medicine, 3) Association of Specialty Professors, 4) Clerkship Directors in Internal Medicine, and 5) Administrators in Internal Medicine. We identified the online Twitter community for this conference through the official hashtag designation established by the AAIM: #APDIM13. Unlike the Twitter learning communities from other scientific meetings, the #APDIM13 hashtag was not created ad-hoc by an unofficial group of conference attendees, but was created and endorsed by the conference organizer (AAIM). Only publicly available tweets and their respective metadata (including author usernames) were collected from the Healthcare Hashtag Project from 28 April to 1 May 20137. The Project provides free “firehose” access to researchers who are investigating the use of Twitter at scientific conferences.\n\nWe performed two separate analyses to quantify Twitter activity based on the number of tweets authored and tweets amplified. In the first analysis, we categorized tweet authors into one of the following groups: 1) faculty, 2) trainee or residency program representative, 3) organization, or 4) other or unidentifiable. Using metadata, we examined the Twitter profile page of each participant of the Twitter community. We categorized participants as “faculty” if his/her profile page indicated s/he was a faculty member at an academic institution. We identified trainees or residency program representatives if their profile page indicated they were a 1) resident, 2) chief resident, or if the username/profile stated they were a residency program (e.g., @ecuimchiefs). We categorized tweets from the AAIM as “organizer”, as they were all participants that represented a third-party organization. We categorized participants as “other” if the profile page was ambiguous or incomplete. We did not perform an internet search of authors whose Twitter profiles were ambiguous because these profiles were deficient in key pieces of information that would have allowed us to identify them correctly (e.g, absence of full names, absence of photograph, and/or unclear location). Finally, we calculated the number/proportion of tweets per category. The greater the proportion of tweets authored, the greater the Twitter activity.\n\nIn the second analysis, we calculated the number of re-tweets per category. Re-tweets are tweets authored by one participant and re-broadcasted (amplified) to a larger Twitter audience by a second participant. We identified re-tweets by the prefix RT within a tweet. Participants whose tweets were re-tweeted the most exhibited high Twitter activity.\n\nWe measured Twitter influence using Google’s algorithm. The PageRank algorithm quantifies individual influence within an online community8–10. It assigns a unitless decimal value to each participant based on three factors: 1) the number of times the incident participant is mentioned in the online community, 2) the number of different participants who mention the incident participant and 3) the PageRank of each participant that mentions the incident participant11–13. For example, if a number of participants mentioned participant A many times, participant A would have a high influence and high PageRank. If a smaller number of participants mentioned participant B, his/her influence would be less than that of participant A14. Each tweet contained the necessary data to determine if the author mentioned another participant.\n\nWe pre-defined “faculty” as individuals from academic institutions who are the most experienced sources of medical information within a Twitter community.\n\nThe data set was downloaded and analyzed using Microsoft Excel 2013. We considered Twitter activity measured as number of tweets and number of tweets re-tweeted, as a continuous variable. We used NodeXL to calculate PageRank as a continuous variable. We used the Fruchterman-Reingold algorithm to develop a directed network map of influence15. Nominal variables included each of the four categories assigned to a participant (faculty, trainee, organization, other/unknown). We used the Chi-square test to compare the nominal variables; t-tests and ANOVA for continuous variable comparisons. We developed a multivariate linear regression model, based on standard least squares, to identify the factors that predicted online influence. JMP Pro 10.1 was used to perform all statistical analyses. We performed a word frequency analysis using NVivo 10. This investigation was exempt from review by the Institutional Review Board because the data set is part of the public domain according to Section 102 of the United States Copyright Act16. To the best of our knowledge, this investigation conforms to STROBE guidelines for observational research and SAMPL guidelines for statistical reporting17,18.\n\n\nResults\n\nOne hundred thirty two participants authored a total of 867 tweets. Common words used in these tweets included: “great”, “residents”, and “meeting” (Figure 1). We identified less than two of every ten participants as a faculty member based on the information from their Twitter profile (19%, 95% CI 13–26%, p < 0.0001). However, the faculty members authored approximately half of all tweets (49%, 95% CI 46–53%, p < 0.0001). Six of every 10 participants did not provide enough information on their Twitter profile to be categorized. There were 261 tweets that were re-tweeted (amplified). Faculty members authored the largest number of amplified tweets (52%, 95% CI 46–58%, p < 0.0001) (Table 1).\n\nThe size of a word represents its relative frequency within the dataset. Word cloud excludes conjunctions, prepositions, articles, specific Twitter usernames (@username), and #APDIM13.\n\nThe mean PageRank for all participants was 0.92 (SD 1.31). Faculty members had the greatest mean PageRank of 1.99 (95% CI 1.53–2.46). This PageRank was statistically greater than that for trainees (1.00, 95% CI 0.47–1.54, p 0.007) and those participants who could not be categorized (0.47, 95% CI 0.20–0.73, p < 0.0001) (Figure 2). Figure 3 shows a pictorial representation of the influence exerted by each participant. The map shows that participants identified as faculty had the largest number of mentions (large density of blue circles/edges).\n\nError bars represent 95% confidence intervals.\n\nGreen triangles (and edges) represent an uncategorized participant, blue circles/edges represent a faculty member, red squares/edges represent organizations and orange diamonds/edges represent trainees or residency programs. Edges are weighted equally. Arrowheads denote the participant who is being mentioned.\n\nWe developed the following multivariate linear regression model:\n\nPageRank = 0.51 + 0.061(number of tweets authored) + 0.067(number of tweets re-tweeted) – 0.25*(1 if category = other) (r2 0.78, p < 0.0001)\n\nIdentifying oneself as a faculty member did not predict the PageRank. The participants that identified themselves as either a trainee or organization did not have higher PageRanks. The model predicted a lower PageRank for those participants who failed to identify themselves or whose identity could not be discerned from their Twitter profile (Table 2).\n\n\nDiscussion\n\nThe two main findings in this investigation are: 1) being an experienced source of medical information has no effect on influence within a Twitter learning community and 2) a large percentage of participants do not provide enough information for one to assess their level of experience.\n\nTwitter learning communities are becoming increasingly popular. Both the American Societies of Clinical Oncology and Nephrology (ASCO and ASN, respectively) have begun yearly Twitter learning communities to accompany their annual scientific meetings4,5. These communities bring together participants of various levels and, effectively, allow each participant to assume the role of both learner and teacher. While “learners” are exposed to a number of “teachers” in these communities, not all participants who assume the role of “teacher” are qualified to do so. Teachers are traditionally considered to have experience regarding the subject matter they teach. These features allow teachers to exert influence over the learners. In our investigation, the most experienced sources of medical information (faculty) exerted the greatest influence in the Twitter community. However, they did so because they had the greatest Twitter activity and not because of their status as faculty members.\n\nInfluence that depends only on Twitter activity and not the experience of the composer of a tweet is concerning. Any participant, regardless of his/her experience, could exert a great influence over the community simply by authoring the most tweets. As a result, learners may be receiving medical information from sources of questionable experience. To our knowledge, there has been no literature to support the idea that participants with the greatest Twitter activity are necessarily the most experienced sources of knowledge.\n\nThe second and equally concerning finding is the ambiguity in Twitter profiles of a large percentage of participants. Uncategorized participants accounted for 58% of all participants in the #APDIM13 community. While over 91% of Twitter users choose to make their profiles publicly visible, fewer seemed to identify their geographic location (75.3%) or place of origin (71.8%)1. Even fewer choose to identify their gender/sex (64.2%)1. Ambiguity in one’s professional status poses a unique challenge in the medical community. Currently, physicians who use Twitter face an “identity dilemma”, which results in incomplete, inaccurate, and often ambiguous Twitter profiles19. Such profiles make it hard for the learner to assess the experience of the participant dispensing information. Moreover, ambiguous profiles are antithetical to the American Medical Association’s (AMA) principles of medical ethics20. Both the AMA and Twitter-savvy physicians advocate “ownership of activity” in social media by avoiding anonymity and accurately stating one’s credentials20,21. Given that we could not identify over half of participants in the #APDIM13 Twitter learning community, it is possible that the current regression model is unable to reveal the predictive power of one’s identity on social media influence.\n\nTwo limitations deserve a special mention. First, we were unsuccessful at identifying those individuals whose Twitter profiles were ambiguous. In this investigation, we classified them as “other” because the vague Twitter profiles did not allow us to identify them with reasonable certainty. Second, we could not include additional variables into our prediction model. The ambiguous profiles did not include information about age, gender, and/or location. Had we included these variables into our multivariate linear equation, we would have produced an unreliable prediction model.\n\nThe greatest strength of this investigation is the method used to measure social media influence. We quantified influence using the number and directionality of mentions within the learning community. Previous studies have used tie strength to measure influence22. Unlike our method, tie strength changes over time, thereby making it difficult to assess one’s influence within a specific learning community. We also used the PageRank algorithm to quantify social media influence. PageRank is considered to be an accurate measurement of influence within social media networks and offers more insight into a person’s influence than simply counting his/her number of followers8–14,23.\n\n\nConclusions\n\nAs the number of Twitter learning communities grow in number and variety, less emphasis will be placed on using social media to exchange medical knowledge. Rather, a greater focus should and will be made towards how to create communities where experienced teachers can 1) be easily identified and 2) have the greatest influence over learners. We must train students to correctly identify experienced sources of information and train those sources to create clear, unambiguous Twitter profiles to allow for easy identification by students. Until then, individuals who consider themselves as experienced educators must actively use Twitter in order to have the greatest influence on learners (Figure 4).\n\n\nData availability\n\nThe data can be downloaded from the Healthcare Hashtag Project website using the hashtag #APDIM13.", "appendix": "Author contributions\n\n\n\nMP and HC were equal contributors to this investigation and are considered co-authors.\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 would like to thank Pooja Desai and Dr. Maria Ferris for their reviews and suggestions in the preparation of this manuscript. We thank Drs. Azeem Elahi, Reed Friend, and Suzanne Kraemer for their social media efforts at East Carolina University.\n\n\nReferences\n\nMislove A, Lehmann S, Ahn YY, et al.: Understanding the Demographics of Twitter Users. Proceedings of the Fifth International AAAI Conference on Weblogs and Social Media. 2011. Accessed on September 10, 2013. Reference Source\n\nShariff A, Fang X, Desai T: Using social media to create a professional network between physician-trainees and the American Society of Nephrology. Adv Chronic Kidney Dis. 2013; 20(4): 357–363. PubMed Abstract \| Publisher Full Text\n\nCheston CC, Flickinger TE, Chisolm MS: Social media use in medical education: a systematic review. Acad Med. 2013; 88(6): 893–901. PubMed Abstract \| Publisher Full Text\n\nDesai T, Shariff A, Shariff A, et al.: Tweeting the meeting: an in-depth analysis of Twitter activity at Kidney Week 2011. PLoS One. 2012; 7(7): e40253. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChaudhry A, Glode M, Gillman M, et al.: Trends in Twitter use by physicians at the american society of clinical oncology annual meeting, 2010 and 2011. J Oncol Pract. 2012; 8(3): 173–178. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMadanick RD, Fleming PS, Kadali R, et al.: Mo1079 Twitter use as a platform for rapid dissemination of informative content from DDW is increasing. Gastroenterology. 2013; 144(5): S571. Reference Source\n\nThe Healthcare Hashtag Project. Accessed August 19, 2011. Reference Source\n\nWilliams S: Is that all there is? A literature review and potential approach to measuring influence in social media. 16th International Public Relations Research Conference. 2013. Accessed September 10, 2013. Reference Source\n\nRubel S: Google’s PageRank is best way to rate online influence. Advert Age. 2008; 79(43): 42. Reference Source\n\nYe S, Wu SF: Measuring message propagation and social influence on Twitter.com. 2010. Accessed September 10, 2013. Reference Source\n\nCha M, Haddadi H, Benevenuto F, et al.: Measuring user influence in Twitter: The Million Follower Fallacy. Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media. 2010; 10. : 10–17. Reference Source\n\nAbdullah IB: Incremental PageRank for Twitter Data Using Hadoop. Master’s Thesis. University of Ediburgh. 2010. Accessed September 10. 2013. Reference Source\n\nPage L, Brin S, Motwani R, et al.: The PageRank Citation Ranking: Bringing Order to the Web. Technical Report - Stanford InfoLab. 1999. Accessed September 10, 2013. Reference Source\n\nLeonhardt D: A Better Way to Measure Twitter Influence. The 6th Floor: Eavesdropping on the Times Magazine. 2011. Accessed on September 10, 2013. Reference Source\n\nKobourov SG: Force-Directed Drawing Algorithms. Accessed September 10, 2013. Reference Source\n\nCopyright Protection not Available for Names, Titles, or Short Phrases Circular 34. 2010. Accessed August 19, 2011. Reference Source\n\nvon Elm E, Altman DG, Egger M, et al.: The strengthening the reporting of observational studies in epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. PLoS Medicine. 2007; 4(10): 1623–1627. Publisher Full Text\n\nLang TA, Altman DG: Basic Statistical Reporting for Articles Published in the Biomedical Journals: The “Statistical Analyses and Methods in the Published Literature” or The “SAMPL Guidelines”. 2013. Accessed October 29, 2013. Reference Source\n\nDeCamp M, Koenig TW, Chisolm MS: Social media and physicians’ online identity crisis. JAMA. 2013; 310(6): 581–582. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChretien KC, Kind T: Social media and clinical care: ethical, professional, and social implications. Circulation. 2013; 127(13): 1413–1421. PubMed Abstract \| Publisher Full Text\n\nDizon DS, Graham D, Thompson MA, et al.: Practical guidance: the use of social media in oncology practice. J Oncol Pract. 2012; 8(5): e114–e124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones JJ, Settle JE, Bond RM, et al.: Inferring tie strength from online directed behavior. PLoS One. 2013; 8(1): e52168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeskovec J: Social Media Analytics: Rich Interactions. 17th Annual Association for Computing Machinery Knowledge, Discovery and Data Mining Conference. 2011. Accessed October 25, 2013. Reference Source" }	[ { "id": "4940", "date": "02 Jun 2014", "name": "Manish Ponda", "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\nSocial media is an increasingly important tool for medical education. This article adds to the necessary body of literature for a greater understanding of its applications.", "responses": [] }, { "id": "4941", "date": "05 Jun 2014", "name": "Paul Phelan", "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\nDesai et al. have reported on the factors contributing to social media influence in a medical community, specifically regarding Twitter use related to an Internal Medicine conference. With a growing presence of social media and online tools in medical education, this represents a relevant and timely study. They demonstrated that faculty members were more likely to have tweets amplified (retweeted) but faculty were not necessarily more influential (as measure using PageRank). Factors influencing influence were number of tweets i,e he/she who shouts the loudest, had the most influence. This has implications for how we get our online material and the quality of the source.Many users (60%) were uncategorized so this undoubtedly took away from the power of the study/conclusions that can be drawn. This is not the fault of the authors however and is a talking point in its own right.Some minor points:Abstract: This portion is confusing to a casual reader: \"Faculty had the greatest influence amongst all participants (mean 1.99, p < 0.0001). Being a faculty member had no predictive effect on influence (β = 0.068, p = 0.6).\"- Regarding PageRank (which many readers will not be familiar with): Does it have a maximum value or is it indefinite? Are RT's included in PageRank? In Results:\"We identified less than two of every ten participants as a faculty member based on the information from their Twitter profile (19%, 95% CI 13–26%, p < 0.0001).\"What does the p value refer to? What is the comparator? This is not immediately obvious to me. Figure 3: Does this add any value?Overall, an interesting & relevant study for medical educators, from authors with a strong track record in this field. It makes us question our sources of online medical material and highlights some issues with twitter use in medical education.", "responses": [] } ]	1	https://f1000research.com/articles/3-120
https://f1000research.com/articles/3-119/v1	28 May 14	{ "type": "Opinion Article", "title": "New forms of checks and balances are needed to improve research integrity", "authors": [ "Elizabeth Iorns", "Christin Chong", "Christin Chong" ], "abstract": "Recent attempts at replicating highly-cited peer-reviewed studies demonstrate that the “reproducibility crisis” is indeed upon us. However, punitive measures against individuals committing research misconduct are neither sufficient nor useful because this is a systemic issue stemming from a lack of positive incentive. As an alternative approach, here we propose a system of checks and balances for the publishing process that involves 1) technical review of methodology by publishers, and 2) incentivizing direct replication of key experimental results. Together, these actions will help restore the self-correcting nature of scientific discovery.", "keywords": [ "The scientific method provides a systematic framework for formulating", "testing and refining hypotheses. By definition", "it requires findings to be reliable so that theories can be refined and scientific progress can occur. Recently", "it has become clear that the scientific method as it is currently being practiced is failing in self-correction", "with multiple studies indicating that more than 70% of surveyed peer-reviewed articles cannot be independently verified1–4. Unfortunately", "instead of focusing on new systems to promote high quality reproducible research", "most resources and attention are focused on trying to police the scientific community by investigating allegations of research misconduct. This approach is destined to fail", "because the problem is systemic and not caused by a few bad players who can be caught and punished. From 1994–2003", "259 cases of misconduct were formally investigated by the Office of Research Integrity5. In contrast", "~480", "000 papers funded by the NIH were published6. It would be impractical and ineffective to investigate why 70% of published findings are irreproducible", "even though ultimately the ability to repeat and build upon prior work is the key component of research integrity that we should care about. Instead", "truly addressing the “reproducibility crisis” requires establishing new checks and balances for the publishing process through 1) technical review of methodology by publishers", "and 2) incentivizing direct replication of key experimental results. If we", "the scientific community", "fail to ensure the quality of the research we produce", "other parties with their own vested interests will step in to police us instead7." ], "content": "Introduction\n\nThe scientific method provides a systematic framework for formulating, testing and refining hypotheses. By definition, it requires findings to be reliable so that theories can be refined and scientific progress can occur. Recently, it has become clear that the scientific method as it is currently being practiced is failing in self-correction, with multiple studies indicating that more than 70% of surveyed peer-reviewed articles cannot be independently verified1–4. Unfortunately, instead of focusing on new systems to promote high quality reproducible research, most resources and attention are focused on trying to police the scientific community by investigating allegations of research misconduct. This approach is destined to fail, because the problem is systemic and not caused by a few bad players who can be caught and punished. From 1994–2003, 259 cases of misconduct were formally investigated by the Office of Research Integrity5. In contrast, ~480,000 papers funded by the NIH were published6. It would be impractical and ineffective to investigate why 70% of published findings are irreproducible, even though ultimately the ability to repeat and build upon prior work is the key component of research integrity that we should care about. Instead, truly addressing the “reproducibility crisis” requires establishing new checks and balances for the publishing process through 1) technical review of methodology by publishers, and 2) incentivizing direct replication of key experimental results. If we, the scientific community, fail to ensure the quality of the research we produce, other parties with their own vested interests will step in to police us instead7.\n\nPublishers are uniquely placed to significantly improve reproducibility because of their inherent need to garner respect from the scientific community. Nature and EMBO are two stand-out examples who are leading the way on ensuring the quality of the research published in their journals. Moreover, current efforts to ensure quality using peer-review alone to weed out irreproducible research are not effective. One reason is that the breadth of technical knowledge that is now required to review a single study is beyond individual scientists. The number of authors per article has increased over the last decade8. In contrast, peer review still relies on two or three peers who are unlikely to be qualified to assess every experimental technique in the study. Nature has implemented an impressive new policy to reduce irreproducibility of its published papers9, and a key aspect to this is employing expert statisticians to review the statistical analysis of papers. Currently, a major limiting factor for implementing technical review is the lack of standardization for methodology design and required controls. Establishing and implementing these standards to ensure the technical quality of the research published in their journals is an effective value-added service that publishers should provide as a separate power in the scientific community. The Resource Identification Initiative (https://www.force11.org/node/4463 date accessed: 2014-04-24) is an example of practical implementation for reporting of materials and methods in a standardized and machine-readable manner. Similar to successful mandates on open access to raw data, journals wield the power to require clear methodology as prerequisite for publication. Further, analogous to open data, the nascent implementation of standardized methodologies will likely yield debates, but lively discussions by the scientific community are useful for policy refinement (http://blogs.plos.org/everyone/2014/03/08/plos-new-data-policy-public-access-data/ date accessed: 2014-04-25).\n\nWhile journals should carry technical review responsibilities, establishing positive incentive structures for reproducible science is necessary to balance the pressure of producing high-profile publications at all costs. Of course, there will always be edge cases where it is not practical to directly replicate findings (for example unpredictable or one-off events like an earthquake), but for the majority of findings it should be possible to directly replicate them. That is, repeat the experiment as-is, while collecting additional information such as “the reliability of the original results across samples, settings, measures, occasions, or instrumentation”10. This is separate from conceptual replication, which is “an attempt to validate the interpretation of the original observation by manipulating or measuring the same conceptual variables using different techniques”10. It is also separate from re-analysis of existing raw data to check for errors in analysis and presentation, but where no new data are obtained. Therefore, directly reproducing experiments is not merely redundant effort, because new data are generated and analyzed to demonstrate the robustness of the original results.\n\nJournals such as F1000Research and PLOS ONE (http://f1000research.com/author-guidelines, http://www.plosone.org/static/publication, date accessed: 2014-03-14) now consider direct replication of original studies, but even a place to publish is not sufficient because there needs to be an effective system to incentivize scientists to conduct replication studies in the first place. The simplest way to conduct replication studies is via fee-for-service technical providers because of their pre-existing methodological expertise and neutral academic involvement (i.e. they are motivated by an operational or a monetary incentive, and thus do not fear retribution from their peers or have the need to accumulate high impact ‘novel’ publications). Similarly, grants specifically designated for research integrity are vital for driving replication (http://www.arnoldfoundation.org/reproducibility-initiative-receives-13m-grant-validate-50-landmark-cancer-studies date accessed: 2014-04-28). These are strategies used by the Reproducibility Initiative (https://www.scienceexchange.com/reproducibility, date accessed: 2014-03-14), and it remains to be proven whether it will be a cost-effective mechanism to conduct direct replications.\n\nThe recent ascent of crowd-sourced post publication peer reviews have identified manuscripts with problematic content, but they remain most active for articles on new techniques that other researchers are eager to replicate for their own experiments (e.g. http://www.ipscell.com/stap-new-data/ date accessed: 2014-04-28 and http://f1000research.com/articles/3-102/v1 date accessed: 2014-05-20). Therefore, positively incentivizing direct replication is necessary for science to become self-correcting again, because no one would selectively publish only their experiments that worked or manipulate their findings knowing that a replication attempt, whether experimental or analytical, would not find the same significant outcome. Scientists would also be more willing to share their raw data and full methodologies before publishing because they want to make sure that their findings are reproducible. Not identifying robust and reproducible research is very costly and impairs our ability to make effective progress against diseases like cancer in which we have already invested billions of dollars. Establishing new checks and balances with existing members of the scientific community such as publishers and fellow scientists is infinitely more preferable than those imposed by outside authorities. And if science progresses by “standing on the shoulders of giants”, it is our duty as scientists to ensure that the “shoulders” are steadfast for our peers.", "appendix": "Author contributions\n\n\n\nE.I and C.C. co-wrote this article.\n\n\nCompeting interests\n\n\n\nElizabeth Iorns is employed at and owns shares of Science Exchange Inc.\n\nChristin Chong has no conflicts of interest to disclose.\n\n\nGrant information\n\n\n\n\nAcknowledgments\n\nWe would like to acknowledge the Reproducibility Initiative board of advisors for their support.\n\n\nReferences\n\nPrinz F, Schlange T, Asadullah K: Believe it or not: how much can we rely on published data on potential drug targets? Nat Rev Drug Discov. 2011; 10(9): 712. PubMed Abstract \| Publisher Full Text\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\nScott S, Kranz JE, Cole J, et al.: Design, power, and interpretation of studies in the standard murine model of ALS. Amyotroph Lateral Scler. 2008; 9(1): 4–15. PubMed Abstract \| Publisher Full Text\n\nPatsopoulos NA, Tatsioni A, Ioannidis JP: Claims of sex differences: an empirical assessment in genetic associations. JAMA. 2007; 298(8): 880–93. PubMed Abstract \| Publisher Full Text\n\nORI Closed Investigations into Misconduct Allegations Involving Research Supported by the Public Health Service: 1994–2003. Lawrence J. Rhoades, Ph.D. Reference Source\n\nBoyack KW, Jordan P: Metrics associated with NIH funding: a high-level view. J Am Med Inform Assoc. 2011; 18(4): 423–431. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMervis J: U.S. science policy. Bill would set new rules for choosing NSF grants. Science. 2013; 340(6132): 534. PubMed Abstract \| Publisher Full Text\n\nPapatheodorou SI, Trikalinos TA, Ioannidis JP: Inflated numbers of authors over time have not been just due to increasing research complexity. J Clin Epidemiol. 2008; 61(6): 546–51. PubMed Abstract \| Publisher Full Text\n\nAnnouncement: Reducing our irreproducibility. Nature. 2013; 496(7446): 398. Publisher Full Text\n\nCollaboration, Open Science, The Reproducibility Project: A Model of Large-Scale Collaboration for Empirical Research on Reproducibility. 2013. Publisher Full Text" }	[ { "id": "4922", "date": "10 Jun 2014", "name": "Andrew D. Chalmers", "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\nImproving reproducibility is a key challenge and topical area in the life sciences. The submitted manuscript provides a well written and interesting commentary on the topic and suggests two key approaches to improve reproducibility, based on technical review and incentivizing replication.I think the paper is suitable for publication, but suggest the authors consider the following comments if they produce a revised version.It is fair to praise Nature and EMBO’s recent efforts, but many scientists would put some blame for current problems on cut down methods sections, driven by space constraints which were/are imposed by some journals such as Nature and EMBO. Standardised methodologies would need to be implemented carefully to avoid stifling scientific progress in developing methods and I suggest this would need to involve scientists as well as publishers? I believe that clearer and longer methods sections are an important and easily achievable way to help improve reproducibility, we have written comments on one small aspect of this, the reporting of antibody use (Helsby MA, Fenn JR and Chalmers AD (2013) Reporting research antibody use: how to increase experimental reproducibility [v2; ref status: indexed, http://f1000r.es/1np ] F1000Research 2013, 2:153 (doi: 10.12688/f1000research.2-153.v2). The authors mention the important RII, but I suggest they could give more prominence to the importance of comprehensive reporting of methods, controls and reagents. It would link directly to their point about better technical review as this is impossible without having well documented methods. The section (2) on different ways to carry out replication could more specifically mention individual scientists trying to replicate findings for their own research, this work is already carried out so involves no additional funding. The key (as mentioned) is then incentivising scientists to publish this work. I wonder what the authors think of initiatives like PubMed commons, aimed at collecting comments on papers, would this provide a format for shorter comments on the ability to reproduce key findings where the scientist concerned might not feel the data warranted a full publication? Is this another useful example of crowd-sourced post publication review?", "responses": [] }, { "id": "4917", "date": "16 Jun 2014", "name": "Ivan Oransky", "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\nThank you for the opportunity to review this article. It makes an important argument in a critical area of inquiry, and deserves publication.I have some specific suggestions for improvement below:\"...punitive measures against individuals committing research misconduct are neither sufficient nor useful because this is a systemic issue stemming from a lack of positive incentive.\"I'd agree that such measures are not sufficient, but what is the evidence that they are not useful? \"From 1994–2003, 259 cases of misconduct were formally investigated by the Office of Research Integrity5. In contrast, ~480,000 papers funded by the NIH were published6. It would be impractical and ineffective to investigate why 70% of published findings are irreproducible, even though ultimately the ability to repeat and build upon prior work is the key component of research integrity that we should care about.\" While it is useful to discuss ORI's limited resources, there are more recent data on their investigations, for example Figure 4 of this paper: http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.1001563. I'd also make it clear that the ORI only has jurisdiction over fabrication, falsification, and plagiarism (FFP) – e.g. scientific misconduct -- and there is no evidence that FFP is responsible for most irreproducibility. So I wouldn't rely on ORI stats for why it's impractical and ineffective to investigate irreproducibility. \"...peer review still relies on two or three peers who are unlikely to be qualified to assess every experimental technique in the study.\"I agree, but can the authors say more about how standardization of methodology design and required controls will solve this problem? \"The recent ascent of crowd-sourced post publication peer reviews have identified manuscripts with problematic content, but they remain most active for articles on new techniques that other researchers are eager to replicate for their own experiments (e.g. http://www.ipscell.com/stap-new-data/ date accessed: 2014-04-28 and http://f1000research.com/articles/3-102/v1 date accessed: 2014-05-20).\"While these two examples demonstrate cases in which post-publication peer reviews are \"on new techniques that other researchers are eager to replicate for their own experiments,\" I'm not sure that's really where post-publication peer review is most active. I would mention PubPeer here, at the very least for context. The authors make a few comments about costs, which are welcome: \"...it remains to be proven whether it will be a cost-effective mechanism to conduct direct replications.\" and \"Not identifying robust and reproducible research is very costly and impairs our ability to make effective progress against diseases like cancer in which we have already invested billions of dollars.\" It would be useful to try to estimate how much replication efforts will cost, and where this funding will come from.", "responses": [] }, { "id": "4918", "date": "18 Jun 2014", "name": "David Soll", "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\nIorns and Chong state in the first paragraph of their Opinion Article that “70% of surveyed peer-reviewed articles cannot be independently verified”. Iorns, who heads the company, Science Exchange, Inc., reported the same statistic in an interview with Jennifer Welsh in Business Insider, 2012. Now she and Christin Chong present a set of recommendations for alleviating this problem. But the way they support their claim that 70% of research is irreproducible is problematic. They base their value primarily on four references 1, 2, 3, 4 that demand scrutiny. These references include three cases on drug effects that are marginal and a fourth on sex differences. Two of the references include data not peer reviewed and authored by individuals from commercial companies1, 2. A third is retrospective and involves the re-evaluation of statistical calculations of the original authors3. Only one, testing the effects of drugs on increased longevity of SOD1G934 mice, provides data that can be assessed4 , and even those data, obtained in an impressive manner, are presented in a review article.There is merit in questioning the reproducibility of studies on marginal drug effects or sex differences, but it seems irresponsible to present, as Iorns and Chong have, a sweeping statement that 70% of all published peer-reviewed articles are irreproducible, even with the qualification of “surveyed” articles. Do these authors really believe that this 70% value applies to studies on signal transduction pathways, the phenotypes of mutants from viruses to bacteria to mammals, the interactions and roles of cytoskeletal molecules, the molecular evolution of species, the functions of molecules in embryogenesis and a vast variety of other biological fields? If Iorns and Chong had limited their commentary to the efficacy of drugs in model systems with marginal effects, they could have made an important and plausible case. But even then they would have had to do a better job referencing their argument. And to bring up the fact that 259 cases of misconduct were investigated by the Public Health Service, followed by their statement “That in contrast ~480,000 papers funded by the NIH were published.”, appears to be an attempt to globalize the problem by insinuation rather than hard supporting data.The suggestion by the authors that publishers should assess the methods and statistics used by third parties is already in place. It is, obviously, the peer review system, and of course it has its problems. But the insinuation is that this process is failing in 70% of cases. Publishers should indeed be more responsible for making sure that reviewers are selected who can really assess whether the methods employed and the statistics applied are valid, especially when marginal effects are claimed. I am sure that all other scientists would whole heartedly agree with that general suggestion. But a vehicle for immediately replicating data in every published paper is extraordinarily impractical, potentially very expensive and not at all necessary in areas of research in which answers are far-more straight forward. And who would foot the bill? The publishers? They are, in almost all cases, for-profit. For replication, they would charge a small fortune. And would scientists spend half of their research funds replicating other scientist’s discoveries. With the radical decrease in funding we are now experiencing, I would not bet on it. Iorns is co-founder of Science Exchange, Inc., a for profit company that charges scientists to have measurements performed in 900 laboratories worldwide that appear to have been recruited to perform experiments for a fee, and a profit, presumably for them and a presumable cut for Science Exchange, Inc. Would Science Exchange, Inc. be the vehicle for such testing?The authors should realize that big discoveries are immediately reproduced by other scientists, to build on those discoveries. Therefore, most scientists are obsessed with the validity of their results. And reproducibility is a tough chore if scientists do not apply the exact same procedures, under the exact same conditions, with the exact same strains and the exact same reagents. Biological systems, from cell cultures to biofilms to biochemical reactions have inherent plasticity and variability, highly responsive to the smallest changes in genetic background, temperature, composition of the atmosphere, trace elements, source of reagents and extracts, and even the quality of double distilled water. But contradictions in the results published by different laboratories have a way of “shaking themselves out”. Most seasoned biologists at the bench know this is the case. Iorns and Chong have made a reasonable case for a limited area of biomedical research that involves searching for small or marginal effects and which involve apparently high noise levels. But they have presented no proof that supports their claim that 70% of all biomedical research is irreproducible, an overstatement which insinuates a significant number of scientists are at worst actively trying to dupe the rest of the scientific world or at best incompetent. By globalizing the problem to a majority of the entire scientific research community in the first paragraph of their commentary, they have sensationalized the targeted problem.", "responses": [] } ]	1	https://f1000research.com/articles/3-119
https://f1000research.com/articles/3-9/v1	13 Jan 14	{ "type": "Research Article", "title": "Effect of chromosome substitution on intrinsic exercise capacity in mice", "authors": [ "Sean M. Courtney", "Michael P. Massett", "Sean M. Courtney" ], "abstract": "Previous research identified a locus on Chromosome 14 as an important regulator of endurance exercise capacity in mice. The aim of this study was to investigate the effect of chromosome substitution on intrinsic exercise capacity and identify quantitative trait loci (QTL) associated with exercise capacity in mice. Mice from a chromosome substitution strain (CSS) derived from A/J and C57Bl/6J (B6), denoted as B6.A14, were used to assess the contribution of Chromosome 14 to intrinsic exercise capacity. All mice performed a graded exercise test to exhaustion to determine exercise capacity expressed as time (min) or work (kg·m). Exercise time and work were significantly greater in B6 mice than B6.A14 and A/J mice, indicating the presence of a QTL on Chromosome 14 for exercise capacity. To localize exercise-related QTL, 155 B6.A14 x B6 F2 mice were generated for linkage analysis. Suggestive QTL for exercise time (57 cM, 1.75 LOD) and work (57 cM, 2.08 LOD) were identified in the entire B6.A14 x B6 F2 cohort. To identify putative sex-specific QTL, male and female F2 cohorts were analyzed separately. In males, a significant QTL for exercise time (55 cM, 2.28 LOD) and a suggestive QTL for work (55 cM, 2.19 LOD) were identified. In the female cohort, no QTL was identified for time, but a suggestive QTL for work was located at 16 cM (1.8 LOD). These data suggest that one or more QTL on Chromosome 14 regulate exercise capacity. The putative sex-specific QTL further suggest that the genetic architecture underlying exercise capacity is different in males and females. Overall, the results of this study support the use of CSS as a model for the genetic analysis of exercise capacity. Future studies should incorporate the full panel of CSS using male and female mice to dissect the genetic basis for differences in exercise capacity.", "keywords": [ "consomic", "genetic", "inbred mice", "treadmill running" ], "content": "Introduction\n\nCardiorespiratory fitness measured during a graded exercise test is inversely related to the relative risk of cardiovascular disease1,2. Results from human cross-sectional, twin, and prospective studies indicate that genetic factors account for 25–65% of the variation in exercise capacity3,4. Because having higher levels of exercise capacity has been shown to be beneficial for reducing the onset of cardiovascular disease, the physiological factors determining exercise capacity have been widely studied5. However, the genetic contribution to exercise capacity is not completely understood. Presently, several candidate genes contributing to improved exercise capacity have been proposed based on genome wide studies6,7, but these genes account for only a small portion of the variability in exercise capacity or training responses8.\n\nSeveral studies have investigated the genetic factors contributing to exercise capacity using inbred rodent models9–13. One common approach has been to screen multiple rodent strains for exercise capacity, followed by quantitative trait loci (QTL) analyses to identify loci linked to exercise capacity. This approach has been used to identify QTL for exercise capacity in rats10 and mice11,13,14. Research from our laboratory previously identified significant and suggestive QTL on several chromosomes that may house candidate genes that influence variation in exercise capacity13,14. These identified regions overlap with other mouse and human QTL, suggesting that these regions and/or genes are conserved among species13. Mouse Chromosome 14 (Chr 14), for example, contained a significant QTL for intrinsic (pre-training) exercise capacity, a significant QTL for exercise capacity after training, and a suggestive QTL for the change in exercise capacity in response to exercise training13. Several linkage markers for maximal oxygen consumption in the sedentary state in humans map to these exercise-related QTL on mouse Chr 1413,15. Therefore, the present study focused on characterizing the role of Chr 14 in regulating intrinsic exercise capacity.\n\nIn the current study we employed a relatively new mouse model, chromosome substitution strains (CSS) to assess the contribution of individual chromosomes to endurance exercise capacity16,17. CSS mice are made by substituting a single chromosome from a donor inbred strain on the genetic background of a host inbred strain (recipient). Therefore phenotypic differences between the recipient or background strain mice and CSS mice support the presence of a QTL on the substituted chromosome for the phenotype being measured. Results from a previous study identified the A/J strain as having low exercise capacity in comparison to the C57BL/6J (B6) strain14. Therefore we chose to use CSS mice based on A/J and B6 inbred strains.\n\nUtilizing this CSS model, the main purposes of the present study were to investigate the effect of chromosome substitution on intrinsic exercise capacity and to identify QTL regulating intrinsic exercise capacity in mice. We hypothesized that chromosome substitution would significantly affect exercise capacity and therefore confirm the importance of Chr 14 in the genetic regulation of intrinsic exercise capacity in mice. Furthermore, we utilized linkage analysis to map QTL on Chr 14 in progeny from a cross between the CSS and host B6 strain.\n\n\nMethods\n\nAll procedures adhered to the established National Institutes of Health guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee at Texas A&M University. Seven week-old inbred mice (A/J, C57BL/6J (B6), and Chr 14 substitution mice (C57BL/6J-Chr 14A/J/NaJ, abbreviated B6.A14)) (n = 12/strain, 6 male and 6 female mice) were purchased from Jackson Laboratory (Bar Harbor, ME.). Upon arrival at Texas A&M, all mice were given one week to acclimatize to their new environment before assessing exercise capacity. A separate group of male B6 mice were crossed with a separate group of female chromosome substitution B6.A14 mice to generate (B6.A14 × B6) F1 mice. The F1 mice were then intercrossed to produce 155 F2 generation mice (67 male and 88 female mice). All mice were housed in standard hanging polycarbonate cages (43 cm long × 21.5 cm wide × 15 cm high) with hardwood chip bedding and allowed food (Standardized Laboratory Rodent Diet) and water ad libitum. Mice were housed 1–5 mice per cage depending on sex and lineage and maintained on a 12 hr light:dark schedule at an ambient temperature of 22–24°C.\n\nAt 8 weeks of age, all mice were familiarized for two days at 9.0 m/min and 10.0 m/min at 10° for 10 minutes to run on a motorized rodent treadmill (Columbus Instruments, Columbus, OH), with an electric grid (160 V, 0–2 mA) at the rear of the treadmill as described previously13,14. Each mouse then completed two graded exercise tests separated by 48 hrs. Mean values for each mouse were used for statistical analyses. For each performance test, the treadmill was started at 9.0 m/min at 0° grade for 9 minutes as a warm-up. The grade was then increased 5° every 9 minutes up to a final grade of 15° and speed was increased 2.5 m/min from a starting speed of 10 m/min every three minutes until exhaustion. Exercise continued until each mouse refused to run, defined as an inability to maintain running speed in spite of repeated contact with the electric grid13,14. At exhaustion, each mouse was immediately removed from the treadmill and returned to its home cage. Exercise capacity was estimated for each animal using time (minutes) and work (kg·m). Work performed (kg·m) or vertical work was calculated as a product of body weight (kg) and vertical distance (meters), where vertical distance = (distance run)(sinθ), where θ is equal to the angle of the treadmill from 0° to 15°13,14.\n\nAt least 24 hours after the last graded exercise test, all mice were anesthetized by intraperitoneal injection of a ketamine (80 mg/kg) - xylazine (5 mg/kg) cocktail. Mice were subsequently euthanized by exsanguination due to removal of the heart and aorta. Heart, gastrocnemius, plantaris, soleus muscle and liver tissue were excised from mice, washed in ice-cold (4°C) saline, blotted dry to remove excess liquid, and snap frozen in liquid nitrogen. DNA was isolated from 25 mg of liver tissue with a DNeasy Blood and Tissue kit (Qiagen Science, Germantown, Maryland) according to the manufacturer’s instructions and quantified using NanoDrop spectrophotometry. Genotyping was performed using competitive allele-specific polymerase chain reaction (PCR) single nucleotide polymorphism (SNP) genotyping (KBiosciences, Hoddesdon, UK)13. All 155 F2 mice were genotyped using 12 SNPs spaced at approximately 5 cM intervals18.\n\nQTL analyses were performed using R/qtl19. One-dimensional scans were performed on the entire F2 cohort with no additional covariates and with sex included as an additive and interactive covariate20. Permutation tests (1,000 repetitions) were used to identify threshold values for logarithm of odds (LOD) scores for each condition (i.e., with or without covariates) and exercise phenotype21. LOD scores were defined as significant if they surpassed the P < 0.05 threshold and suggestive if they surpassed the P < 0.63 threshold. If suggestive or significant QTL were identified using sex as interactive covariate, then one-dimensional scans were performed on male and female mice separately to identify potential sex-specific QTL. A two-dimensional scan also was performed on the entire F2 cohort to identify additive or interacting QTL on Chr 14. QTL confidence intervals were determined using the 1.5 LOD support interval19.\n\nAll data are represented as mean ± SE. Statistical significance for phenotype comparisons was denoted by P < 0.05. Two-way analysis of variance was used to determine the effect of sex and strain on exercise capacity, which is defined as time (minutes), or work (kg·m) (JMP 9.0, SAS, Cary, NC). If significant main effects were found for strain, Dunnett’s post hoc test was used to determine significant strain differences compared with B6. If significant main effects were found for sex, t-tests were used to identify sex differences within each strain. Comparisons among parental strains and F2 offspring and across genotypes for allelic effects were made using one way analysis of variance (strain or genotype) followed by Tukey’s post-hoc analysis. T-tests were used to identify sex differences in F2 offspring.\n\n\nResults\n\nInbred strains. Exercise capacity, defined as mean run time during two graded exercise treadmill tests, for inbred and CSS mice is shown in Figure 1. Exercise times in A/J and B6.A14 mice were significantly less (P < 0.0001) than that in B6 mice. A significant effect of sex also was identified in all strains (A/J, B6.A14, and B6). For each strain, female mice ran significantly longer than male mice from the same strain (Figure 1A). When exercise capacity was expressed as work, A/J and B6.A14 strains were significantly different from B6 (P < 0.0001), with mice from both strains performing less work than B6 mice (Figure 1B). In contrast to exercise time, there was no significant main effect for sex (P = 0.1) on exercise capacity defined as work. Significant differences among the strains were primarily limited to differences in exercise phenotypes. Body mass was significantly less in B6.A14 mice compared with B6 (P < 0.0008) (Table 1). There were no significant differences in absolute tissue mass among A/J, B6, and B6.A14 strains (Table 1). Within each strain, body mass was significantly lower in females compared to males. Accordingly, tissue masses were lower in female mice compared to male mice from the same strain (Table 1).\n\nExercise capacity was assessed using a graded exercise test and expressed as (A) time in minutes or (B) work in kg·m. Values are expressed as mean ± SE. n = 6 mice per group. P < 0.05 compared to B6; †P < 0.05 compared to females.\n\nValues are mean ± SE. n = 6 per group, except n = 5 for B6 males. HM:BM, heart mass-to-body mass ratio; GM:BM, gastrocnemius mass-to-body mass ratio; SM:BM, soleus mass-to-body mass ratio; PM:BM, plantaris mass-to-body mass ratio.\n\nP < 0.05 significant main effect for strain compared with C57BL/6J.\n\n†P < 0.05 significantly different from male mice of same strain.\n\n(B6.A14 × B6) F2 mice. The sex-specific distributions for exercise time and work in F2 mice are shown in Figure 2. Both time and work varied significantly between male and female F2 mice (Table 2). On average, female mice ran approximately 2.5 min longer than male mice, whereas work was not significantly different between males and females. Body mass was approximately 5 g higher in male mice compared with females (P < 0.0001), which likely accounts for the similar levels of work performed. Similar to body mass, tissue masses were significantly smaller in female F2 mice compared to their male counterparts.\n\nFrequency distribution for (A) time and (B) work in male and female (B6.A14 × B6) F2 mice. All F2 mice (n = 155) performed a graded exercise test to exhaustion to assess exercise capacity. Mice assorted into 1min/0.25 kg·m buckets. n = 67 for males and n = 88 for females.\n\nValues are mean ± SE. HM:BM, heart mass-to-body mass ratio; GM:BM, gastrocnemius mass-to-body mass ratio; SM:BM, soleus mass-to-body mass ratio; PM:BM, plantaris mass-to-body mass ratio.\n\nP < 0.05 significantly different from male mice.\n\nRelative to the progenitor strains, eight week old F2 mice ran an average of 30.8 ± 0.1 min, which was significantly longer (P < 0.0001) than B6.A14 (28.5 ± 0.2 min) and not different from B6 (31.0 ± 0.1 min) mice. F2 mice also performed significantly more work (1.69 ± 0.02 kg·m, P < 0.0001) than B6 (1.39 ± 0.04 kg·m) and B6.A14 (0.95 ± 0.02 kg·m) strains. Body mass also was compared across strains and generations. F2 mice had average body mass of 22.1 ± 0.2 g, which was significantly greater than (P < 0.0001) the progenitor B6 (20.7 ± 0.8 g), and B6.A14 (18.9 ± 0.6 g) strains, respectively.\n\nQTL analysis. Significant differences in exercise capacity between B6.A14 and B6 strains indicate the presence of a QTL on Chr 14. To fine-map the QTL, F2 mice were generated from B6.A14 and B6 strains. Single chromosome-wide scans for time and work are shown in Figure 3. Suggestive QTL for time (LOD = 1.75, P = 0.131) and work (LOD = 2.08, P = 0.063) with no covariates were identified on Chr 14. When sex was included as an interacting covariate, a significant QTL was identified at 56 cM for time (LOD = 3.8, P = 0.048) (Figure 3A). A suggestive QTL for work (LOD = 3.69, P = 0.07) was identified at the same location (Figure 3B). Because significant and suggestive QTL were identified using sex as an interacting covariate, chromosome-wide scans were performed on male and female mice separately. In male mice, a significant QTL for time (LOD = 2.28, 1.5 LOD = 49.0 – 58.9 cM, P = 0.049) and a suggestive QTL for work (LOD = 2.19, 1.5 LOD = 38.0 – 58.9 cM, P = 0.056) were identified at 55 cM. In female mice, no QTL were identified for time Figure 4A). However, a suggestive QTL for work was identified at 16 cM (LOD = 1.8, P = 0.106) (Figure 4B). The two-QTL analyses for time showed limited evidence for additive QTL at 0 cM and 58 cM (LOD = 2.74, P = 0.19) on Chr 14. No significant additive or interacting QTL were identified for work. QTL scans also were performed for all physical characteristics and no significant QTL were identified.\n\nQTL analyses on mouse Chromosome 14 for intrinsic capacity expressed as time (min, A) and work (kg·m, B) in 155 (B6.A14 × B6) F2 mice. Three analyses were performed for each phenotype: 1) with no covariates, 2) with sex as an additive covariate, and 3) with sex as an interactive covariate. For time, significant (P = 0.05) logarithm of odds (LOD) thresholds are 2.22 with no covariates, 2.12 with sex as an additive covariate, and 3.78 with sex as an interactive covariate. For work, significant (P = 0.05) LOD thresholds are 2.22 with no covariates, 2.27 with sex as an additive covariate, and 3.95 with sex as an interactive covariate. LOD thresholds were determined using 1000 permutations. Chromosome-wide scans and permutation analyses were performed using R/qtl.\n\nQTL analyses for the effect of sex on intrinsic exercise capacity in male and female (B6.A14 × B6) F2 mice expressed as time (A) and work (B). Single chromosome-wide scans for time (in min) and work (in kg•m) were performed separately on male and female F2 mice. Dashed lines in the upper graph represent the suggestive (0.82, P = 0.63) and significant (2.21, P = 0.05) logarithm of odds (LOD) thresholds for time in males. Dashed lines in the lower graph represent the suggestive (0.82, P = 0.63) and significant (2.25, P = 0.05) LOD thresholds for work in males. In females, suggestive and significant LOD thresholds for time were 0.84 (P = 0.63) and 2.15 (P = 0.05), respectively; and for work 0.81 (P = 0.63) and 2.08 (P = 0.05), respectively. LOD thresholds were determined using 1000 permutations. Chromosome-wide scans and permutation analyses were performed using R/qtl.\n\nThe allelic effects for suggestive and significant QTL are shown in Table 3. In the entire F2 cohort, heterozygous mice had the highest average exercise time and work. For both phenotypes there was no significant difference between homozygous A and B groups. A similar pattern was observed for time and work in the male F2 cohort. In this group, mice with parental genotypes had significantly lower exercise time than mice carrying the heterozygous genotype (Table 3). In the female F2 cohort, work was significantly higher in homozygous A mice compared with homozygous B mice. Heterozygous female F2 mice had an intermediate phenotype.\n\nPosition, location of peak marker in cM; Marker, SNP marker closest the LOD peak; Genotype, genotype at peak marker; A, homozygous for A/J allele; B, homozygous for B6 allele; H, heterozygous; p-value, p-value from means comparison using Tukey post-hoc analysis with specific allelic comparisons indicated (e.g., A vs. B), significant p-values are indicated by.\n\n\nDiscussion\n\nThe purpose of the current study was to determine the role of mouse Chr 14 in the genetic regulation of exercise capacity and to fine map this chromosome to identify QTL for exercise capacity. Significant differences in exercise time and work were observed between inbred B6 mice and mice carrying Chr 14 from the A/J strain on the B6 background (B6.A14). These differences suggest the presence of one or more QTL on Chr 14 underlying variation in exercise capacity. Utilizing a (B6.A14 × B6) F2 population, suggestive QTL for exercise time and work were localized to a position of ~58 cM on Chr 14. Further analysis revealed putative sex-specific QTL for exercise time and work. QTL identified in the male cohort was similar to that in the entire F2 cohort, but the suggestive QTL for work identified in the female F2 mapped to an alternative position. Collectively, these data suggest that one or more genes on Chr 14 contribute to variation in exercise capacity and that the genetic architecture for exercise-related traits might be different in males and females. Given the complexity of the trait, genome-wide mapping strategies should be employed to identify additional QTL underlying the variation in exercise capacity.\n\nB6 and A/J strains show significant phenotypic differences across many traits16,22–25. We, and others14,26,27 have demonstrated that exercise capacity assessed by treadmill running is one of these traits. Although testing protocols varied, A/J mice repeatedly show low exercise capacity, having running times approximately 60% or less than that of B6 mice. In the current study, A/J mice ran 10 minutes less than B6 mice and performed only 25% of the work of B6 mice during a graded exercise test (Figure 1). These observations replicated our previous finding that A/J mice had the lowest exercise capacity among 34 strains tested14. Although B6 were in the lowest third of that survey, their run time was about 60% higher than A/J mice. This disparity in exercise capacity between inbred strains suggests that genetic variation contributes to these phenotypic differences.\n\nChromosome substitution strains were developed to facilitate genetic analysis of complex traits by partitioning the genome into individual chromosomes16. Phenotypic differences between a CSS and the background strain suggest the presence of at least one QTL on the substituted chromosome. To begin to identify the genetic factors contributing to variation in exercise capacity, mice from a chromosome substitution strain based on A/J and B6 strains were used. In the current study we focused on Chr 14 because we had previously identified several exercise-related QTL on this chromosome13. B6.A14 mice had significantly lower exercise capacity expressed as time or work compared with B6 mice (Figure 1). Exercise time was 2.5 minutes less in B6.A14 mice, which corresponds to 24% of the difference between parental A/J and B6 strains (10.6 min). The difference in work was 0.44 kg·m, which is approximately 43% of the difference between parental strains (1.03 kg·m). Although the substituted chromosome shifted the phenotype toward the donor strain, the effect of chromosome substitution on exercise capacity was less than expected. Based on previous CSS surveys, chromosome substitution can produce phenotypic effects of 75% or more of the difference between parental strains16,23,24. Nevertheless, the significant difference between B6 and B6.A14 for exercise time and work suggest the presence of one or more QTL on Chr 14 for exercise capacity.\n\nTo localize the QTL on Chr 14, linkage analysis was performed in F2 mice from B6 and B6.A14 strains. Suggestive QTL were identified for both time and work at 58 cM (Figure 3). The 1.5 LOD interval for each of these QTL spanned nearly the entire chromosome, so these QTL overlapped with previously reported QTL for pre-training and post-training work13. However, the peak markers for pre-training work (4 cM) and post-training work (26 cM) QTL localize to positions distant from the QTL identified in the current study and likely represent different QTL. Further analysis using sex as an interactive covariate provided evidence for sex-specific QTL; therefore, male and female F2 cohorts were analyzed separately. Significant and suggestive QTL for time and work, respectively, were identified in the male cohort only and were similar to those identified in the entire cohort. Conversely, there was less evidence for exercise-related QTL on Chr 14 in female mice. This is somewhat surprising given the differences between B6 and B6.A14 female mice were comparable to those in male mice (Figure 1). However, the peak marker for the suggestive QTL for work in the female F2 cohort is in close proximity to a syntenic human region linked to maximal oxygen consumption in the sedentary state in the HERITAGE Family Study15.\n\nOne potential explanation for the limited evidence for exercise QTL in the F2 cohort is that the number of animals was insufficient for detecting multiple QTL with small effects. However, the number of mice included in the entire F2 cohort or each sex-specific cohort is comparable to most intercross populations utilizing a CSS and B6 strains and should have been sufficient to detect at least 1 QTL16,24,25. Similar to the current study, Burrage et al. were also unable to localize QTL in CSS × B6 intercross populations for several traits showing significant differences between parental CSS and B6 mice25. They concluded that multiple QTL with opposing effects might be present on individual chromosomes and that congenic strains might be more advantageous for QTL detection and mapping than larger intercross populations. Alternatively, a close inspection of the allelic effects for each exercise QTL suggests that alleles derived from the A/J stain contribute to increasing exercise capacity (Table 3). This was most evident in the female F2 cohort. The suggestive QTL for work identified in this population mapped to a position (16 cM) that was different from that observed in the entire F2 and male-only cohorts. In females, mice homozygous for the parental A allele performed significantly greater work that mice homozygous for the parental B allele. Heterozygous mice were intermediate and not significantly different from either parental genotyping suggesting an additive inheritance pattern with the A allele conferring increasing exercise capacity. In the full F2 and male-only cohorts, there was no significant difference between mice homozygous for the parental genotypes and heterozygous mice had the highest exercise capacity. Thus, at some locations A and B alleles can interact to elicit a phenotype greater that either parental genotype.\n\nCollectively, these data support the use of CSS as a model for the genetic analysis of exercise capacity. They also provide evidence that genetic factors on Chr 14 contribute to the variation in exercise capacity. Based on the complexity of the exercise phenotype, a survey of the complete C57BL/6J-ChrA/J/NaJ CSS panel will likely identify multiple chromosomes of interest and potential QTL for exercise capacity. Furthermore, the sex-dependent differences in exercise capacity and the putative sex-specific QTL imply that the genetic architecture underlying exercise capacity might be different between males and females. Thus, any such survey should be conducted in male and female mice to elucidate the potential genotype by sex interaction underlying differences in exercise capacity between males and females. Once strong candidate genes are identified, the link between exercise capacity and cardiorespiratory fitness, and the mechanistic basis for diseases associated with low cardiorespiratory fitness can be explored.\n\n\nData availability\n\nFigshare: Effect of chromosome 14 substitution on intrinsic exercise capacity in mice: R/qtl linkage analysis and phenotype data, http://dx.doi.org/10.6084/m9.figshare.89358128.", "appendix": "Author contributions\n\n\n\nSC designed and conducted the study, analyzed the data, and wrote the manuscript. SC critically revised the manuscript and agreed to its publication. MM designed the study, analyzed the data, and wrote the manuscript. MM critically revised the manuscript and agreed to its publication.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by NIH grant HL-085918 (MPM) and the Sydney and J.L. Huffines Institute for Sports Medicine and Human Performance (SC).\n\n\nAcknowledgements\n\nThe authors would like to thank Jean Kovar for management of the breeding colony for this study.\n\n\nReferences\n\nKokkinos P, Myers J, Faselis C, et al.: Exercise capacity and mortality in older men: a 20–year follow-up study. Circulation. 2010; 122(8): 790–797. PubMed Abstract \| Publisher Full Text\n\nMyers J, Prakash M, Froelicher V, et al.: Exercise capacity and mortality among men referred for exercise testing. N Engl J Med. 2002; 346(11): 793–801. PubMed Abstract \| Publisher Full Text\n\nBouchard C, Daw EW, Rice T, et al.: Familial resemblance for VO2max in the sedentary state: the HERITAGE family study. Med Sci Sports Exerc. 1998; 30(2): 252–258. PubMed Abstract \| Publisher Full Text\n\nBouchard C, Rankinen T, Timmons JA: Genomics and genetics in the biology of adaptation to exercise. Compr Physiol. 2011; 1(3): 1603–1648. PubMed Abstract \| Publisher Full Text\n\nBooth FW, Roberts CK, Laye MJ: Lack of exercise is a major cause of chronic diseases. Compr Physiol. 2012; 2(2): 1143–1211. PubMed Abstract \| Publisher Full Text\n\nBouchard C, Sarzynski MA, Rice TK, et al.: Genomic predictors of maximal O2 uptake response to standardized exercise training programs. J Appl Physiol. 2011; 110(5): 1160–1170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTimmons JA, Knudsen S, Rankinen T, et al.: Using molecular classification to predict gains in maximal aerobic capacity following endurance exercise training in humans. J Appl Physiol. 2010; 108(6): 1487–1496. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBouchard C: Overcoming barriers to progress in exercise genomics. Exerc Sport Sci Rev. 2011; 39(4): 212–217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarbato JC, Koch LG, Darvish A, et al.: Spectrum of aerobic endurance running performance in eleven inbred strains of rats. J Appl Physiol. 1998; 85(2): 530–536. PubMed Abstract\n\nWays JA, Cicila GT, Garrett MR, et al.: A genome scan for Loci associated with aerobic running capacity in rats. Genomics. 2002; 80(1): 13–20. PubMed Abstract \| Publisher Full Text\n\nLightfoot JT, Turner MJ, Knab AK, et al.: Quantitative trait loci associated with maximal exercise endurance in mice. J Appl Physiol. 2007; 103(1): 105–110. PubMed Abstract \| Publisher Full Text\n\nMassett MP, Berk BC: Strain-dependent differences in responses to exercise training in inbred and hybrid mice. Am J Physiol Regul Integr Comp Physiol. 2005; 288(4): R1006–R1013. PubMed Abstract \| Publisher Full Text\n\nMassett MP, Fan R, Berk BC: Quantitative trait loci for exercise training responses in FVB/NJ and C57BL/6J mice. Physiol Genomics. 2009; 40(1): 15–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCourtney SM, Massett MP: Identification of exercise capacity QTL using association mapping in inbred mice. Physiol Genomics. 2012; 44(19): 948–955. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBouchard C, Rankinen T, Chagnon YC, et al.: Genomic scan for maximal oxygen uptake and its response to training in the HERITAGE Family Study. J Appl Physiol. 2000; 88(2): 551–559. PubMed Abstract\n\nSinger JB, Hill AE, Burrage LC, et al.: Genetic dissection of complex traits with chromosome substitution strains of mice. Science. 2004; 304(5669): 445–448. PubMed Abstract \| Publisher Full Text\n\nNadeau JH, Singer JB, Matin A, et al.: Analysing complex genetic traits with chromosome substitution strains. Nat Genet. 2000; 24(3): 221–225. PubMed Abstract \| Publisher Full Text\n\nPetkov PM, Ding Y, Cassell MA, et al.: An efficient SNP system for mouse genome scanning and elucidating strain relationships. Genome Res. 2004; 14(9): 1806–1811. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBroman KW: Review of statistical methods for QTL mapping in experimental crosses. Lab Anim (NY). 2001; 30(7): 44–52. PubMed Abstract\n\nKorstanje R, Li R, Howard T, et al.: Influence of sex and diet on quantitative trait loci for HDL cholesterol levels in an SM/J by NZB/BlNJ intercross population. J Lipid Res. 2004; 45(5): 881–888. PubMed Abstract \| Publisher Full Text\n\nChurchill GA, Doerge RW: Empirical threshold values for quantitative trait mapping. Genetics. 1994; 138(3): 963–971. PubMed Abstract \| Free Full Text\n\nHoit BD, Kiatchoosakun S, Restivo J, et al.: Naturally occurring variation in cardiovascular traits among inbred mouse strains. Genomics. 2002; 79(5): 679–685. PubMed Abstract\n\nShao H, Burrage LC, Sinasac DS, et al.: Genetic architecture of complex traits: large phenotypic effects and pervasive epistasis. Proc Natl Acad Sci U S A. 2008; 105(50): 19910–19914. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStylianou IM, Tsaih SW, DiPetrillo K, et al.: Complex genetic architecture revealed by analysis of high-density lipoprotein cholesterol in chromosome substitution strains and F2 crosses. Genetics. 2006; 174(2): 999–1007. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurrage LC, Baskin-Hill AE, Sinasac DS, et al.: Genetic resistance to diet-induced obesity in chromosome substitution strains of mice. Mamm Genome. 2010; 21(3–4): 115–129. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLightfoot JT, Turner MJ, Debate KA, et al.: Interstrain variation in murine aerobic capacity. Med Sci Sports Exerc. 2001; 33(12): 2053–2057. PubMed Abstract \| Publisher Full Text\n\nNadeau JH, Burrage LC, Restivo J, et al.: Pleiotropy, homeostasis, and functional networks based on assays of cardiovascular traits in genetically randomized populations. Genome Res. 2003; 13(9): 2082–2091. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCourtney SM, Massett MP: Effect of chromosome 14 substitution on intrinsic exercise capacity in mice: R/qtl linkage analysis and phenotype data F1000Research. figshare 2014. Data Source" }	[ { "id": "3576", "date": "31 Mar 2014", "name": "Michael Turner", "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 is very well written regarding a strong, well thought out research design. The research is cutting edge and moves the knowledge regarding possible genetic factors influencing intrinsic exercise capacity in active mice. Chromosome 14 is an important participant in this phenotype and the authors do an excellent job utilizing chromosome substitution to continue pursuing their research interests. Minor comments: The authors point out in the first paragraph of the Results that body mass was less in the B6.A14 mice compared to the B6 mice. Could the authors perform a simple regression to assess the influence of differing body mass on intrinsic exercise capacity as a potential confounding factor to their findings?\n\nThe third sentence in the second paragraph of the Results appears to conflict with the previous statement regarding differences in work with sex in the F2 generation. If I'm reading these two sentences incorrectly I would ask a clearer statement be made to help the reader understand the sex-related difference in work with these F2 mice. In the last few sentences in the fourth paragraph of the Discussion the authors present an interesting discussion regarding the different QTLs on chromosome 14 that may be playing a role in the male vs. female F2 mice and intrinsic exercise capacity. More discussion would add to the major findings and future directions for readers. This is a difficult research issue and a bit more discussion would direct the readers towards an appreciation of this position. Possibly a better transition to the argument provided in the following paragraph would assist in the interpretation of the authors' findings.", "responses": [ { "c_id": "780", "date": "23 Apr 2014", "name": "Michael Massett", "role": "Author Response", "response": "The authors point out in the first paragraph of the Results that body mass was less in the B6.A14 mice compared to the B6 mice. Could the authors perform a simple regression to assess the influence of differing body mass on intrinsic exercise capacity as a potential confounding factor to their findings? We thank the Reviewer for this suggestion. We performed a linear regression to determine the relation between body mass and exercise capacity expressed as time. We analyzed each of the inbred strains separately as well as the F2 population. For each of these groups, exercise time was significantly negatively correlated with body mass. In general, body mass explained 50% to 74% of the variance in exercise time in the inbred strains. In the F2 population, the variance in exercise time explained by body mass was approximately 54%; however, this percentage decreased when the population was divided by sex. In males this percentage was 22 % and in females, less than 2%. We will include the correlation data in the revised manuscript to identify both the direction and magnitude of this relation. The third sentence in the second paragraph of the Results appears to conflict with the previous statement regarding differences in work with sex in the F generation. If I'm reading these two sentences incorrectly I would ask a clearer statement be made to help the reader understand the sex-related difference in work with these F2 mice. We thank the Reviewer for pointing out this confusing description. Simply stated, work is determined as the product of exercise capacity and body mass. Female mice ran longer/farther than male mice, but weighed less; therefore work was approximately equal between male and female F2 mice. We will edit this description in the revised manuscript. In the last few sentences in the fourth paragraph of the Discussion the authors present an interesting discussion regarding the different QTLs on chromosome 14 that may be playing a role in the male vs. female F mice and intrinsic exercise capacity. More discussion would add to the major findings and future directions for readers. This is a difficult research issue and a bit more discussion would direct the readers towards an appreciation of this position. Possibly a better transition to the argument provided in the following paragraph would assist in the interpretation of the authors' findings. We are grateful for the Reviewer’s appreciation of the complexity of the phenotype and the number of genetic and environmental factors that influence this phenotype. We will expand our discussion of some of these genetic factors in the revised manuscript." } ] }, { "id": "4461", "date": "11 Apr 2014", "name": "Reuben Howden", "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 give the impression that a higher, genetically controlled, exercise capacity should equal lower disease risk based on previous work. However, I wonder if a genetic predisposition to a higher exercise capacity is more or less important that habitual exercise itself. It would be interesting to investigate disease risk in inbred mice using training programs of differing volume/intensity so the effect of environmental factors (e.g. habitual exercise) can be assessed in a controlled environment. Habitual exercise and disease risk reductions may be achievable without developing an exceptional exercise capacity. Prof. Edward Howley’s recent comments about recommending vigorous exercise, needed for high exercise capacity, from the end of his lecture1 may be pertinent here. Assessment of exercise capacity by treadmill running - this may be an age old question, but it remains an important one; how do the authors know they successfully assessed exercise capacity in these strains and that differences in treadmill running time were not influenced by individual strain motivation to continue exercise? While this is not a study design flaw, it could change the question asked about factors influencing sedentary behavior and chronic disease risk in human populations. As the authors point out, a limitation of their study is the small F2 cohort. While the reported cohort size may be concordant with previous work, that does not mean there were sufficient meiotic events to fine map their previously identified QTL on Chr14.\n\nIt is difficult to determine from this article if the Chr14 QTL was reduced in the current study. Perhaps the authors could provide more information about that.\n\nDid the use of CSS increase the authors confidence about specific candidate genes identified previously?2 What potential candidate genes are located within the present (refined?) Chr14 QTL? I agree that congenic strains might be useful for identifying which regions within the reported Chr14 QTL, although because of the likely complexity of factors influencing exercise capacity, demonstrating a phenotypic influence of more specific genomic regions or individual candidate genes may be challenging", "responses": [ { "c_id": "779", "date": "23 Apr 2014", "name": "Michael Massett", "role": "Author Response", "response": "“The authors give the impression that a higher, genetically controlled, exercise capacity should equal lower disease risk based on previous work. However, I wonder if a genetic predisposition to a higher exercise capacity is more or less important that habitual exercise itself. It would be interesting to investigate disease risk in inbred mice using training programs of differing volume/intensity so the effect of environmental factors (e.g. habitual exercise) can be assessed in a controlled environment. Habitual exercise and disease risk reductions may be achievable without developing an exceptional exercise capacity. Prof. Edward Howley’s recent comments about recommending vigorous exercise, needed for high exercise capacity, from the end of his lecture may be pertinent here.” The Reviewer raises several good points in this comment. First, our goal in the Introduction was to remind the readers of the well-documented inverse relationship between cardiorespiratory fitness and morbidity and mortality in humans. This relationship appears to plateau at 10 METs, suggesting that having a fitness level above 10 METs does not confer significant additional protection (Kokkinos et al., 2008). Therefore, we agree with the Reviewer that developing high levels of exercise capacity and/or fitness are not required to achieve beneficial effects.Regarding the question of fitness versus physical activity, the Reviewer is correct in pointing out that increasing physical activity can reduce the risk of all-cause mortality and CVD and that some benefit may be gained by increasing activity without improving fitness (maximal oxygen uptake). However, Williams (2001) did point out that cardiorespiratory fitness and physical activity should be considered independently as predictors of disease risk and that the relative risk of heart disease in the most fit group is nearly half that of the most active group. These data imply that both fitness and activity are important factors in reducing the risk of cardiovascular disease. Determining the genetic basis for each phenotype could yield important information regarding the beneficial effects of exercise. “Assessment of exercise capacity by treadmill running - this may be an age old question, but it remains an important one; how do the authors know they successfully assessed exercise capacity in these strains and that differences in treadmill running time were not influenced by individual strain motivation to continue exercise? While this is not a study design flaw, it could change the question asked about factors influencing sedentary behavior and chronic disease risk in human populations.” In the current study, exercise performance was assessed during a graded exercise test, which was stopped when mice exhibited pre-defined behaviors that we, and others, have defined as exhaustion. As pointed out by Booth, Laye and Spangenburg (2010), in ideal settings, exhaustion would be defined by a number of physiological markers. The large number of mice required for genetic experiments exceeded our capacity to assess maximal oxygen consumption and/or blood lactate levels at the time of testing. However, Desai et al. (1997) showed that an exercise test similar to the one utilized in the current study elicited heart rates near 750 beats per minute and a respiratory exchange ratio at or above 1.0 in mice from several inbred strains. We cannot assess an animal’s motivation to run on the treadmill so we do not know how much this might have contributed to our results. But, as demonstrated by Lerman et al. (2002) the willingness to run voluntarily is likely determined by different genetic factors than those underlying the variation in intrinsic exercise capacity in the untrained state. This latter phenotype is the focus of the current study. “As the authors point out, a limitation of their study is the small F cohort. While the reported cohort size may be concordant with previous work, that does not mean there were sufficient meiotic events to fine map their previously identified QTL on Chr14.” The number of mice utilized in this study was sufficient to detect at least one QTL explaining 5% of the variance in exercise capacity with a power of 50% (Belknap, 2003).Increasing the number of mice would likely have allowed us to detect more QTL with smaller effects. However, as the reviewer points out, an experimental model with a greater number of recombination events, such as an advanced intercross line based on A and B6 strains, the hybrid mouse diversity panel, or mice from the collaborative cross, would probably have yielded a greater resolution for our QTL interval. “It is difficult to determine from this article if the Chr14 QTL was reduced in the current study. Perhaps the authors could provide more information about that.” The QTL for pre-training work on Chromosome 14 identified in our previous study was located at 4 cM with a 95% CI of 0-38 cM. The suggestive QTL for time and work identified in the current study were located near 58 cM and the 95% CI were very large. Although the QTL in the current study did overlap with our previously identified QTL, it is very likely that the QTL identified in the current study are different from our previous report. A brief statement regarding this issue is located in the fourth paragraph of the Discussion section. “Did the use of CSS increase the authors confidence about specific candidate genes identified previously? What potential candidate genes are located within the present (refined?) Chr14 QTL?” Because there was no overlap between the current and previous QTL, the use of the CSS model did not increase our confidence regarding specific candidate genes in the Chromosome 14 QTL region. However, given the complexity of the exercise capacity phenotype, performing a genome scan using a complete set of CSS might be a logical step toward identifying chromosomes of interest and potential candidate genes. These chromosomes could then be targeted for more detailed analyses. “I agree that congenic strains might be useful for identifying which regions within the reported Chr14 QTL, although because of the likely complexity of factors influencing exercise capacity, demonstrating a phenotypic influence of more specific genomic regions or individual candidate genes may be challenging” We agree with the Reviewer that identifying candidate genes for exercise and exercise-related traits is challenging. However, the availability of resources for genetic and genomic studies continues to increase, which should facilitate discovery of genes and gene networks that regulate variation in exercise capacity, fitness, and potentially diseases associated with low levels of fitness." } ] } ]	1	https://f1000research.com/articles/3-9
https://f1000research.com/articles/3-118/v1	28 May 14	{ "type": "Case Report", "title": "Lung transplantation in an intensive care patient with pulmonary alveolar microlithiasis - a case report", "authors": [ "Bülent Güçyetmez", "Aylin Ogan", "Aylin Çimet Ayyıldız", "Berrin Yalçın Güder", "Walter Klepetko", "Aylin Ogan", "Aylin Çimet Ayyıldız", "Berrin Yalçın Güder", "Walter Klepetko" ], "abstract": "Introduction: Pulmonary alveolar microlithiasis (PAM) is an autosomal recessive disease characterized by the deposition of phosphate and calcium in the alveoli. The disease progresses asymptomatically until later stages. When it becomes symptomatic, lung transplantations performed before the onset of right heart failure may improve life expectancy and quality. Here we present a case report concerning the very first Turkish PAM patient to have undergone lung transplantation surgery.Patient information: A 52 year-old female, Caucasian patient, already diagnosed with PAM in infancy, was admitted to the intensive care unit, diagnosed with pneumonia and hospitalized for 20 days. We decided to refer the patient to a specialized center for lung transplantation. Bilateral lung transplantation was performed in Vienna 14 months later and no recurrence was observed during the first postoperative year.Conclusion: Bilateral lung transplantation may improve both the life expectancy and the quality of life of PAM diagnosed patients with severe respiratory failure who do not suffer from right heart failure. The risk of recurrence should not be considered as a justifying reason to avoid transplantation as a treatment method.", "keywords": [ "Lung transplantation", "pulmonary alveolar microlithiasis" ], "content": "Introduction\n\nPulmonary alveolar microlithiasis (PAM) was first described by Harbitz in 19181. This rare disease which progresses with calcium and phosphate deposition in the alveolar space is an autosomal recessive disorder caused by the SLC34A2 gene mutation2,3. Radiological images reveal typically bilateral, diffuse and symmetrical sandstorm-like widespread radiopaque micronodules4. Turkey is the country with the highest PAM prevalence (16.3%), followed by Italy and USA5,6. The only known treatment is lung transplantation performed before the onset of right heart failure. No recurrence has been reported after transplantation7,8. Here we present the case report of the first Turkish patient followed-up in the intensive care unit (ICU) with the diagnosis of PAM, who needed ventilator support at the time of discharge from the ICU and received lung transplantation in Austria.\n\n\nPatient information\n\nA 52 year-old female patient, with a family history of PAM, was diagnosed with the same disease when she was 10 years old and received no treatment or intervention until 2012. The patient affected by PAM presented with tachypnea, exertional dyspnea and fatigue to the Emergency Department and she was admitted to the ICU in 2012 with suspicion of pneumonia. At the time of ICU admission, she was conscious, cooperative and the initial vital signs were SpO2: 57% (spontaneous respiration under 5lt/min mask O2 support), pulse rate: 127/mn: blood pressure: 126/65mmHg, body temperature: 37°C, C-reactive protein (CRP): 10.83 and leucocyte: 9600/mm3. The arterial blood gas values (10lt/min mask O2) were detected as pH:7.45 PaO2: 53.5mmHg PaCO2: 34mmHg SaO2: 86.5% HCO3: 24.9 mmol/L base excess: 0.3 mmol/L Na: 133 mmol/L K: 4.8 mmol/L Cl: 106 mmol/L Ca: 1.09 mmol/L lactate: 1.3 mmol/L. The patient was given non-invasive mechanical ventilation (NIMV) support with positive end-expiratory pressure (PEEP): 10cmH2O and inspiratory pressure (IP): 22cmH2O FiO2: 60%. The chest X-ray and thoracic computed tomography (CT) taken at the ICU admission revealed bilateral, diffuse involvement (sandstorm) and decreased aeration areas of both lungs (Figure 1, Figure 2). The infection markers (body temperature, leucocyte, CRP) of the patient receiving NIMV support during the 20 days of ICU hospitalization improved after the 15th day. The daily respiratory parameters (respiratory rate, PaO2, PaO2:FiO2 ratio, SpO2), infection markers and the administered medications are shown in Figure 3 and Figure 4. Since the patient still had the consistent need of NIMV support despite the improvement of the laboratory values, and no changes were detected in the radiological images, she was evaluated together with Pulmonary Diseases and Cardiology departments. Following thoracic CT, echocardiography (mild pulmonary hypertension, EF 60%) and respiratory function tests (Forced Expiratory Volume in 1 second, FEV1:0.51L Forced Vital Capacity, FVC:0.54L FEV1/FVC:0.94), the patient was discharged from the ICU on the 20th day and referred to a specialized lung transplantation center. No extrapulmonary involvement was observed by Positron Emission Tomography (PET). The patient underwent bilateral lung transplantation in Vienna 14 months after initial admission (she was under oxygen and NIMV support during these 14 months). She was followed-up during the first seven postoperative days in the ICU and discharged on the 21st day from the hospital. On the postoperative 6th month, the values of the patient, having no need for oxygen or NIMV support, were FEV1:2.21 FVC:2.26 FEV1/FVC:98%. The results of chest X-rays taken on the postoperative 1st month and 1st year, the thoracic CT scans taken on the postoperative 6th month and 1st year and the arterial blood gas under room air on the postoperative 6th month of the patient administered mycophenolic acid 760mg/day, tacrolimus 0.5mg/day, and prednisolone 5mg/day medication are demonstrated in Figure 1, Figure 2 and Figure 3 respectively.\n\na; ICU admission (typical chest image of PAM) b; 6th month after transplantation c; 1st year after transplantation.\n\na; ICU admission (bilateral sandstorm image) b; 1st month after transplantation c; 1st year after transplantation.\n\n\nDiscussion\n\nThe etiology, epidemiology, clinical findings and typical radiological images related to PAM disease have been almost completely documented. The aim here is to discuss the advantages of lung transplantation as a treatment option for PAM.\n\nIt has been described that PAM is an autosomal recessively inherited disorder related to genetic factors2,3. Although PAM is rarely observed in infants9, the clinical findings and the radiological changes advance progressively over-time; micronodular structure (sandstorm) develops due to the deposition of calcium and phosphate, aeration areas decrease, fibrosis increases and hypoxemia occurs. Patients presenting to the hospital with these clinical findings are generally over the age of 40 and no administered treatments result in full recovery. Systemic corticosteroids, calcium-chelating agents and bronchoalveoler lavage (BAL) are palliative solutions10. Ozçelik et al. have described the positive effects of the long term use of sodium etidronate which is effective by inhibiting the hydroxyapatite microcrystal formation in pediatric patients11. However, there are also some studies showing that the sodium etidronate treatment is ineffective7.\n\nThe priority for these patients admitted to the ICU should be to seek solutions for recovering hypoxemia. It is observed that patients presenting highly decreased aeration areas have already undergone many treatment methods11. In Figure 4 we show that the patient was administered sodium etidronate, methylprednisolone and sildenafil in the ICU. Besides, no improvement was observed in the radiological images or respiratory parameters despite the oxygen and NIMV support and the regress in the infection markers. Even at the time of discharge from the ICU, the patient was under NIMV support, and was oxygen dependent with a peripheral saturation of 90%, PaO2 90–95mmHg and with PaO2/FiO2 ratio under 150. The patient clinically worsened with infection; after the regression of the infection, she could only return to her previous hypoxemic condition despite NIMV support.\n\nThe issue that intensivists should discuss is whether a patient with a hypoxemic condition should be discharged with the recommendation of oxygen and NIMV support or whether the transplant choice should be brought forward.\n\nAlthough Turkey is the country with the highest PAM prevalence, we have not encountered any published reports on transplanted cases upon the diagnosis of PAM in Turkey. Lung transplantation is a treatment option for PAM patients and is recommended in severe cases of oxygen dependent respiratory failure before the onset of right heart failure7,8,10. Bonette et al. recommended bilateral lung transplantations instead of unilateral transplants on the grounds of resistant shunt development in a lung transplantation series of 14 cases, one of which was caused by PAM12. Besides, no recurrence was detected in the presented cases or series7,8,10,13,14. Moon et al. reported the case of a patient who survived for 15 postoperative years with no recurrence after the lung transplantation was performed after PAM diagnosis8. In their series, Shigemura et al. reported no recurrence and only two cases of postoperative major bleeding after bilateral lung transplantation15. Furthermore, they reported a significant increase in the FEV1 and FVC levels of the patients and considered bilateral lung transplantation in PAM cases as a successful and reliable treatment method.\n\nThus, considering lung transplantation as a treatment method in PAM cases instead of medication or artificial respiratory support treatments, both of which are known to be ineffective, seems rational. In our case, we discussed the lung transplantation indication of this PAM-diagnosed patient to offer her a chance of full recovery. The differences observed between the preoperative and postoperative periods demonstrated the importance of performing bilateral lung transplantation in such cases regardless of the recurrence risk. Not only did the radiological follow-up for one year show no recurrence; but also the patient receiving immunosuppressive treatment no longer needed oxygen or artificial respiratory support systems.\n\n\nConclusion\n\nBilateral lung transplantation may improve both the life expectancy and quality of PAM-diagnosed patients with severe respiratory failure who do not suffer from right heart failure. The risk of recurrence should not be considered as a valid reason to eliminate transplantation option as a treatment method.\n\n\nKey messages\n\n- All the other treatment methods in PAM are palliative except for transplantation.\n\n- A PAM-diagnosed patient being followed-up in the ICU due to severe respiratory failure needs oxygen and NIMV support even at the time of discharge.\n\n- Intensivists should discuss the transplantation option in cooperation with the pulmonologists, cardiologists and transplantation team; they should also have an active role in the management of the PAM-diagnosed patients after their discharge.\n\n- The risk of recurrence should not be considered as a justifying reason to avoid the transplantation option.\n\n\nConsent\n\nWritten informed consent for publication of this case report and accompanying figures was obtained from the patient.", "appendix": "Author contributions\n\n\n\nBG was involved in the initial writing of the manuscript. AO, AÇA and BYG were primarily involved in the care of our patient. WK provided intellectual contributions to the content of the manuscript as well as editorial assistance. All authors have read and approved the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interest.\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 Idil Cakir for her editorial contribution.\n\n\nReferences\n\nHarbitz F: Extensive calcification of the lungs as a distinct disease. Arch Intern Med. 1918; 21(1): 139–146. Publisher Full Text\n\nDogan OT, Ozsahin SL, Gul E, et al.: A frame-shift mutation in the SLC34A2 gene in three patients with pulmonary alveolar microlithiasis in a inbred family. Intern Med. 2010; 49(1): 45–9. PubMed Abstract \| Publisher Full Text\n\nOzbudak IH, Bsşsorgun CI, Ozbilim G, et al.: Pulmonary alveolar microlithiasis with homozygous c.316G > C (p.G106R) mutation: a case report. Turk Patoloji Derg. 2012; 28(3): 282–5. PubMed Abstract \| Publisher Full Text\n\nGasparetto EL, Tazoniero P, Escuissato DL, et al.: Pulmonary alveolar microlithiasis presenting with crazy-paving pattern on high resolution CT. Br J Radiol. 2004; 77(923): 974–6. PubMed Abstract \| Publisher Full Text\n\nMariotta S, Ricci A, Papale M, et al.: Pulmonary alveolar microlithiasis: report on 576 cases published in the literature. Sarcoidosis Vasc Diffuse Lung Dis. 2004; 21(3): 173–81. PubMed Abstract\n\nCastellana G, Lamorgese V: Pulmonary alveolar microlithiasis. World cases and review of the literature. Respiration. 2003; 70(5): 549–55. PubMed Abstract \| Publisher Full Text\n\nJönsson AL, Simonsen U, Hilberg O, et al.: Pulmonary alveolar microlithiasis: two case reports and review of the literature. Eur Respir Rev. 2012; 21(125): 249–56. PubMed Abstract \| Publisher Full Text\n\nMoon E, Tsuang W, Bonnette P, et al.: Lung transplantation outcomes for pulmonary alveolar microlithiasis: a limited case series. 2009. Publisher Full Text\n\nYin J, Shen K: Images in clinical medicine. Pulmonary alveolar microlithiasis in a child. N Engl J Med. 2011; 364(22): e49. PubMed Abstract \| Publisher Full Text\n\nSamano MN, Waisberg DR, Canzian M, et al.: Lung transplantation for pulmonary alveolar microlithiasis: a case report. Clinics (Sao Paulo). 2010; 65(2): 233–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOzcelik U, Yalcin E, Ariyurek M, et al.: Long-term results of disodium etidronate treatment in pulmonary alveolar microlithiasis. Pediatr Pulmonol. 2010; 45(5): 514–7. PubMed Abstract \| Publisher Full Text\n\nBonnette P, Bisson A, el Kadi NB, et al.: Bilateral single lung transplantation. Complications and results in 14 patients. Eur J Cardiothorac Surg. 1992; 6(10): 550–4. PubMed Abstract \| Publisher Full Text\n\nStamatis G, Zerkowski HR, Doetsch N, et al.: Sequential bilateral lung transplantation for pulmonary alveolar microlithiasis. Ann Thorac Surg. 1993; 56(4): 972–5. PubMed Abstract \| Publisher Full Text\n\nJackson KB, Modry DL, Halenar J, et al.: Single lung transplantation for pulmonary alveolar microlithiasis. J Heart Lung Transplant. 2001; 20: 226. PubMed Abstract \| Publisher Full Text\n\nShigemura N, Bermudez C, Hattler BG, et al.: Lung transplantation for pulmonary microlithiasis. J Thorac Cardiovasc Surg. 2010; 139(3): e50–2. PubMed Abstract \| Publisher Full Text" }	[ { "id": "5369", "date": "07 Jul 2014", "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\n\nThis case report represents the state of the art in the care of these patients very well. This patient was admitted to the hospital with life threatening respiratory failure. The underlying cause was diagnosed and treated. An excellent discharge plan was created and implemented. The patient was referred for lung transplantation, and underwent the procedure with an apparently good functional result. The duration of follow up is relatively short.", "responses": [] }, { "id": "5518", "date": "24 Jul 2014", "name": "Johan Groeneveld", "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\nWell documented case report. I wonder whether patients with mild cor pulmonale would also qualify for this life saving procedure, since this may ameliorate further remodeling of the heart. Another thing that bothers me is whether this patient was on selective decontamination of the digestive tract during postoperative ICU stay or not? The documentation of pulmonary infections and their (appropriate) antibiotic treatment is somewhat lacking. This is an important issue for lung transplant, immune suppressed patients. Did this patient harbor multi-resistant microorganisms prior to transplant?", "responses": [] } ]	1	https://f1000research.com/articles/3-118
https://f1000research.com/articles/1-27/v1	11 Oct 12	{ "type": "Research Article", "title": "Evolution of alarm cues: a role for kin selection?", "authors": [ "Denis Meuthen", "Sebastian A Baldauf", "Timo Thünken", "Sebastian A Baldauf", "Timo Thünken" ], "abstract": "The evolution of alarm signals has puzzled evolutionary ecologists for decades. This is particularly true for alarm cues (‘Schreckstoff’) which are present in many fishes. They are passively released through injuries and signal the presence of a predator. Although the benefits for conspecific receivers are obvious (they can adjust their behavior to avoid predation), those for the senders are not which is, however, a necessary requirement for the evolution of alarm signals. Several hypotheses relying on potential direct benefits for the senders have been suggested. Alarm cues might attract secondary predators which in turn might increase the escape probability of the sender. A primary immune enhancing role was suggested as well. An alternative explanation is based on Hamilton’s inclusive fitness theory stating that individuals can indirectly increase their fitness by increasing the survival of genetically related individuals (‘kin selection theory’). If related individuals preferentially benefit from alarm signals, for instance by being more receptive to kin-alarm cues, senders could increase their inclusive fitness. Here, we investigate whether individuals of the cichlid fish Pelvicachromis taeniatus respond differentially to alarm cues derived from kin and non-kin. P. taeniatus possesses alarm cues and is known to adjust its behavior when exposed to alarm cues. We measured the change in activity after the addition of alarm cues (derived from kin- and non-kin) relative to a control treatment. Reduced activity is a widespread behavioral adaptation to reduce predation risk in prey organisms. Fish of the alarm cue treatments significantly reduced their activity relative to control fish. However, fish did not respond differentially to alarm cues derived from kin and non-kin suggesting that potential inclusive fitness benefits are not mediated by responses specific to individual alarm cues. We discuss alternative mechanisms such as kin-related grouping and mating preferences potentially leading to kin-biased alarm cue perception.", "keywords": [ "Please note that the refereeing status of this article was changed from “indexed” to “[v1", "ref status: approved with reservations 2]”." ], "content": "Introduction\n\nAlarm signals are signals that are emitted by individuals in the presence of predators. These signals may not only divert predator attention or discourage it to attack but also alert conspecifics1. Alarm signals can be transmitted via different communication pathways: alarm calls are widespread among social terrestrial animals (rodents2, birds3 and primates4). Other animal species use alarm pheromones for the same purpose (insects5,6 or mice7). In aquatic environments, chemical alarm signals are widespread especially due to the large number of water-soluble compounds8 (insects9, crustaceans10, asteroids11,12, gastropods13,14, amphibians15,16 and fishes17). Signaling is usually costly for the individual sender18 and often appears to primarily benefit the receivers. However, to evolve and be maintained by natural selection, the fitness benefits for the signaling individual must override its costs.\n\nThe evolution of alarm cues in fishes is a particularly puzzling example of alarm signaling whose evolution has intrigued evolutionary ecologists since the 1960’s19,20 until today21. Alarm cues, also termed ‘Schreckstoff’’22,23 are located in specialized epidermal ‘club cells’24–26. They are released following injuries by predators, and widespread among different fish taxa including cichlids25,27. Signal-receiving conspecifics benefit because the presence of alarm cues reliably indicates high predation risk, thus allowing them to respond to predator presence in order to increase survival28,29. However, the benefits to the signaler are unclear because it often may not survive a predatory attack. Furthermore, alarm cue production is energetically demanding30.\n\nSeveral hypotheses have been proposed to explain the phenomenon of alarm cues in fishes31,32. First, alarm cues might benefit the survival of the producing individual. In this context, it has been proposed that alarm pheromones may function to attract secondary predators33. During the following interference between competing predators prey might be able to escape. Mathis et al. showed that pike Esox lucius and predatory diving beetles were indeed attracted by minnow alarm cues33. In the presence of such secondary predators (pike), escape probabilities of minnows were actually increased34. Accordingly, the signaling individual can increase its own fitness by producing alarm cues, provided that it survives an initial predator attack.\n\nOther authors proposed that the alarm function is a mere by-product and instead suggested the primary function of alarm cues to be anti-pathogenic agents35 or promote the healing of injuries36. A recent study supported this hypothesis; club cell production was unrelated to predation risk but stimulated by skin penetrating pathogens and parasites37. Moreover they found that UV radiation also affected club cell production, providing evidence for a general immune function of alarm cues.\n\nA further hypothesis relying on Hamilton’s inclusive fitness theory predicts that individuals can increase their fitness indirectly by supporting the survival of genetically related individuals38, coined ‘kin selection’ by Maynard-Smith39. Generally, costly communication is facilitated when interacting individuals are related40. While kin selection has been suggested to play an important role in the evolution of mammalian alarm calls2,41–43 (but see44), its significance for the evolution of alarm cues in fishes has received little attention30. Indirect benefits can be gained when kin particularly benefit from alarm cue production. This might be reached when individuals are able to discriminate between the alarm signals of different senders45. Higher sensitivity to kin-alarm cues may result in an improved response to predation, and thus higher survival of individuals related to the sender which in turn may increase the indirect fitness of the sender.\n\nThe aim of the present experiment was to test whether the cichlid fish Pelvicachromis taeniatus discriminates between alarm cues produced by kin and non-kin. P. taeniatus is a socially monogamous small cave-breeder with biparental brood care46 which inhabits streams in Western Africa47. Previous studies revealed that this species possesses alarm cues, recognizes conspecific alarm cues and adjusts its behavior in the presence of alarm cues (Meuthen et al., submitted, unpublished data). Furthermore, P. taeniatus is capable of kin recognition48–50 which is most likely based on chemical cues51–53. In the experiment we measured the change in activity in individual P. taeniatus after the addition of alarm cues derived from kin and non-kin, respectively.\n\n\nMaterial and methods\n\nThis study conforms to the Association for the Study of Animal Behaviour’s Guidelines for the Use of Animals in Research and was carried out according to the German laws for experimentation with animals (§ 8 Abs. 1 TierSchG, V.m. § 2 Abs. 1.1 TierSchZustV NW 26.9.1989). No additional licences were required.\n\nWe conducted an experiment using female F2 progeny of the cichlid species Pelvicachromis taeniatus, whose ancestors (F0) were collected from the Moliwe river near Limbe, Cameroon (04°04’ N, 09°16’ E). Female P. taeniatus were used exclusively due to their consistent activity levels54. Prior to experiments, fish were kept in mixed-sex 50 × 30 × 30 cm (L × W × H) stock tanks at densities up to 20 individuals and were fed daily with frozen invertebrate larvae ad libitum. These tanks were illuminated in a 12:12 h light:dark cycle; water temperature was kept at 25 ± 1°C.\n\nDuring experiments, we manipulated predation risk in three treatments by exposing individual fish to one of the following extracts: (1) Alarm cues derived from kin (from familiar and unfamiliar siblings); (2) Alarm cues derived from unfamiliar, unrelated conspecifics (non-kin); (3) Distilled water to control for disturbance effects upon introduction. We produced alarm cues from 26 donor cichlids which were previously starved for two days to exclude any effects caused by the individual’s selective diet. Each alarm cue consisted of a male and a female cichlid to control for sex effects. Fish were anaesthetized with a blow to the head and afterwards euthanized by cervical dislocation in accordance to § 4 of the German animal welfare act (BGB l. I S. 1207, 1313). They were then placed into a mortar and grinded with a pestle. This procedure, imitating a lethal predation event, ruptured cells and thus allowed alarm pheromones to be released. The homogenate was diluted with distilled water, passed through filter floss and frozen in 1 ml aliquots at -20°C until use. The final concentration each fish was exposed to during trials was 3.6 mg/l donor wet fish weight. Likewise, we prepared 1 ml aliquots of pure distilled water for control experiments.\n\nTrials were run in 30 × 20 × 20 cm tanks which were supplemented with a 0.5 × 0.5 cm grey plastic tube leading below the water level at the middle of their short side. This duct allowed the direct addition of chemical cues into the tanks while minimizing fish disturbance. Furthermore, experimental tanks were surrounded on all sides (except the top) with white polystyrene to prevent fish agitation by neighboring fish or the experimenter. A video camera (QuickCam 9000, Logitech, China) viewing the tanks from the top enabled recording of fish behavior for evaluation. Tanks were filled with substrate-treated water54; individual fish were then introduced and acclimatized for 1 h, this period is referred to as the pre-stimulus phase from now on. Experimental stimuli were thereafter temperature-adjusted to tank conditions and introduced at the point of 1 h 15min. Subsequently, fish behavior was recorded for another hour. Afterwards, the experimental subjects were sized accurate to the nearest millimeter and weighed accurate to one milligram on an electrical precision scale (LC 2215, Sartorius, Germany). Between trials, tanks were cleaned with 3% hydrogen peroxide and then rinsed with tap water to remove remaining olfactory traces55,56. Furthermore, experimental stimuli assigned to individual tanks were alternated between trials.\n\nIn total, we tested 51 individuals from 8 families; extracts from the same donor fish were used throughout different treatments and thus represented – based on the family identity of the focal fish – as either related or unrelated conspecifics. Also, individuals from the same family were evenly distributed among the three extracts. Because in two cases the fish exhibited no activity during the pre-stimulus phase, we excluded them from analysis. Hence, the final sample size consisted of 49 individuals; 12 received alarm cues from familiar siblings, 9 from unfamiliar siblings, 14 from unrelated fish and 14 individuals were exposed to the control stimulus. Because sibling familiarity did not significantly affect focal fish activity (familiar vs. unfamiliar kin: χ2 = 0.090, p = 0.764), their activity scores were pooled to represent 21 fish receiving alarm cues derived from kin.\n\nWe evaluated fish activity by tracking its movement during 1 h pre-stimulus phase and the following 1 h post-stimulus phase with animal tracking software (Biobserve Viewer2, St. Augustin, Germany). Subsequently, we assigned an activity index to each fish by calculating the difference between the distances covered during the two experimental phases. Activity indices did not deviate significantly from normal distribution according to the Kolmogorov-Smirnov test (Lillie.test, R library “nortest”), thus we applied linear-mixed effect models (LME, R library “nlme”) for analysis. All test fish were only used once but to account for the repeated use families, we entered “family identity” as a random factor. All results were based on likelihood ratio tests (LRT); hence degrees of freedom always differed by one.\n\n\nResults\n\nThe change in activity of female P. taeniatus was significantly affected by the treatment (χ2 = 10.057, p = 0.007, Figure 1). Activity indices of both alarm-cue treatment groups (kin/non-kin) were significantly different from those of the water-control group (Kin vs. water: χ2 = 8.346, p = 0.004; Non-kin vs. water: χ2 = 8.693, p = 0.003, Figure 1). Whereas fish of the control group showed on average an increase of 0.67 m in covered distance during the post-stimulus period, fish of both alarm cue treatments showed reduced activity in the post-stimulus phase (on average 0.31 m less compared to the pre-stimulus phase). However, fish did not respond differentially to alarm cues derived from related and unrelated individuals (χ2 = 0.233, p = 0.630).\n\nActivity indices were calculated by subtracting the distance covered during the 1 h prestimulus phase from the following 1 h poststimulus period. Asterisks above the bars indicate ** p < 0.01; ns p > 0.6.\n\n\nDiscussion\n\nGenerally, activity of female P. taeniatus was affected by the presence of conspecific alarm cues. Fish of the alarm cue treatment significantly decreased their activity relative to control fish. However, P. taeniatus did not discriminate between alarm cues derived from kin or non-kin in terms of activity changes.\n\nThese results are in accordance to numerous studies showing that the presence of conspecific alarm cues decreases prey activity in general57,58. Reduced activity concurrently decreases prey conspicuousness, which is an effective strategy against visual predators and enhances prey survival58. Furthermore, our results add to an earlier study, showing that males of P. taeniatus reduce territorial aggression in the presence of conspecific alarm cues (Meuthen et al., submitted, unpublished data). Thus both sexes of P. taeniatus are capable of recognizing alarm cues and adjust their behavior accordingly.\n\nAgainst the expectations of the kin selection hypothesis, fish did not discriminate between kin and non-kin alarm cues. This is also surprising because theoretical models proposed that the evolution of communication in general is driven by the interaction between siblings40 and kin selection has been suggested to play an important role in the evolution of different alarm signals such as alarm calls2,41–43 but see44. Furthermore, individual discrimination of alarm signals is predicted to be highly beneficial45. However, lack of discrimination does not necessarily mean lack of recognition59. Thus, further studies in different contexts are required to confirm the present results. The results are in accordance with the “direct benefits” hypotheses32, including those hypotheses highlighting that alarm cues have primarily evolved as an immune enhancing mechanism and that the alarm signaling function is a by-product37. They are also in accordance with the theory that chemical signals have originally evolved from compounds without a communicative function8.\n\nOn a molecular level, the missing discrimination may result from a lack of kin-related information within alarm cues. Although to date, some studies were able to identify single chemical components of alarm cues such as Hypoxanthine-3-N-oxide60,61 or the glycosaminoglycan chondroitin62, the full suite of alarm cue components still remains to be identified62. The results of behavioral experiments have suggested that substantial individual variation in alarm cues exists. Sender body condition was determined to be an important factor affecting alarm cue response63. These results were, however, attributed to differences in club cell quantity30 because fish display graded responses based on alarm cue concentration64. Unidentified qualitative effects could, however, also be responsible for the altered response, especially because not all behaviors were specific to sender condition – whereas changes in activity and shoaling density were specific to sender condition, aggression remained unaffected65.\n\nKin discrimination in alarm cues might simply not have evolved in P. taeniatus due to the lack of necessity. The increase in fitness due to kin discrimination in alarm cues might not be sufficient to cause the necessary impact on natural selection, causing kin-related information to evolve in alarm cues. Many fishes are capable of identifying kin and live in kin-shaped groups66. This is also true for P. taeniatus which lives the first weeks of hatching in family group guarded by their parents46. Afterwards they live in shoals without defined territories throughout the juvenile stage47. Laboratory experiments showed that they prefer to shoal with familiar kin (Thünken et al., unpublished data). Furthermore, P. taeniatus preferentially mate with kin47 which is most likely also true for the natural population67. Accordingly, throughout their life the social environment of P. taeniatus is probably largely kin-structured in nature. In this case, individuals receiving the information transmitted by alarm cues are most likely kin. As a consequence, signaling individuals may increase their inclusive fitness.\n\nIn conclusion, our study found no evidence for differential response to kin derived alarm cues. However, under natural conditions behavioral mechanisms may lead to kin-biased alarm cue perception. Thus, kin selection potentially plays a role in alarm signaling in our model system. Still, further research is required determining the direct fitness benefits and costs for the signaling individual as well as the benefits for the receivers which are fundamental parameters to understand the evolution and maintenance of alarm cues.", "appendix": "Author contributions\n\n\n\nTT and DM conceived the study. DM, SAB and TT designed the experiments. DM carried out the research. DM, SAB and TT analysed the data. DM and TT wrote the paper. All authors had read and improved 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\nThis research was funded by the Deutsche Forschungsgemeinschaft (TH 1615/1-1).\n\n\nAcknowledgements\n\nWe are grateful to the Bakker research group for discussion.\n\n\nReferences\n\nSmith RJF: Evolution of alarm signals: Role of benefits of retaining group members or territorial neighbors. Am Nat. 1986; 128(4): 604–610. Reference Source\n\nSherman PW: Nepotism and the evolution of alarm calls. Science. 1977; 197(4310): 1246–1253. PubMed Abstract \| Publisher Full Text\n\nKlump GM, Shalter MD: Acoustic behavior of birds and mammals in the predator context. 1. Factors affecting the structure of alarm signals. 2. The functional significance and evolution of alarm signals. Z Tierpsychol. 1984; 66(3): 189–226. Publisher Full Text\n\nMacedonia JM, Evans CS: Essay on Contemporary Issues in Ethology: Variation among mammalian alarm call systems and the problem of meaning in animal signals. Ethology. 1993; 93(3): 177–197. Publisher Full Text\n\nBlum MS: Alarm pheromones. Annu Rev Entomol. 1969; 14: 57–80. Publisher Full Text\n\nKunert G, Otto S, Rose USR, et al.: Alarm pheromone mediates production of winged dispersal morphs in aphids. Ecol Lett. 2005; 8: 596–603. Publisher Full Text\n\nRottman SJ, Snowdon CT: Demonstration and analysis of an alarm pheromone in mice. J Comp Physiol Psychol. 1972; 81(3): 483–490. PubMed Abstract \| Publisher Full Text\n\nSteiger S, Schmitt T, Schaefer HM, et al.: The origin and dynamic evolution of chemical information transfer. Proc Biol Sci. 2011; 278(1708): 970–979. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSih A: Antipredator responses and the perception of danger by mosquito larvae. Ecology. 1986; 67(2): 434–441. Publisher Full Text\n\nLaforsch C, Beccara L, Tollrian R, et al.: Inducible defenses: The relevance of chemical alarm cues in Daphnia. Limnol Oceanogr. 2006; 51(3): 1466–1472. Publisher Full Text\n\nParker DA, Shulman MJ: Avoiding predation: Alarm responses of Caribbean sea-urchins to simulated predation on conspecific and heterospecific sea-urchins. Mar Biol. 1986; 93(2): 201–208. Publisher Full Text\n\nLawrence JM: A chemical alarm response in Pycnopodia helianthoides (Echinodermata: Asteroidea). Mar Behav Physiol. 1991; 19(1): 39–44. Publisher Full Text\n\nSleeper HL, Paul VJ, Fenical W, et al.: Alarm pheromones from the marine opisthobranch Navanax inermis. J Chem Ecol. 1980; 6(1): 57–70. Publisher Full Text\n\nKempendorff W: Über das Fluchtphänomen und die Chemoreception von Helisoma (Taphius) nigricans. Arch Molluskenkd. 1942.\n\nHews DK, Blaustein AR: An investigation of the alarm response in Bufo boreas and Rana cascadae tadpoles. Behav Neural Biol. 1985; 43(1): 47–57. PubMed Abstract \| Publisher Full Text\n\nKats LB, Petranka JW, Sih A, et al.: Antipredator defenses and the persistence of amphibian larvae with fishes. Ecology. 1988; 69(6): 1865–1870. Publisher Full Text\n\nWisenden BD: Olfactory assessment of predation risk in the aquatic environment. Philos Trans R Soc Lond B Biol Sci. 2000; 355(1401): 1205–1208. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHughes AL: Evolution of adaptive phenotypic traits without positive Darwinian selection. Heredity (Edinb). 2012; 108(4): 347–353. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilliams GC: Measurement of consociation among fishes and comments on the evolution of schooling. Publications of the Museum, Michigan State University Biological Series. 1964; 2: 349–384. Reference Source\n\nWilliams GC: Natural selection: domains, levels, and challenges. (Oxford University Press, 1992). Reference Source\n\nChivers DP, Brown GE, Ferrari MCO, et al.: Chemical ecology in aquatic systems. eds C. Brönmark & L. A. Hansson (Oxford University Press) 2012; 312. Reference Source\n\nvon Frisch K: Zur Psychologie des Fisch-Schwarmes. Naturwissenschaften. 1938; 26(37): 601–606. Publisher Full Text\n\nvon Frisch K: Über einen Schreckstoff der Fischhaut und seine biologische Bedeutung. Z Vgl Physiol. 1942; 29(1–2): 46–145. Publisher Full Text\n\nBarreto RE, Barbosa A, Giassi ACC, et al.: The ‘club’ cell and behavioural and physiological responses to chemical alarm cues in the Nile tilapia. Mar Freshw Behav Physiol. 2010; 43(1): 75–81. Publisher Full Text\n\nPfeiffer W: The distribution of fright reaction and alarm substance cells in fishes. Copeia. 1977; 1977(4): 653–665. Reference Source\n\nKristensen EA, Closs GP: Anti-predator response of naive and experienced common bully to chemical alarm cues. J Fish Biol. 2004; 64(3): 643–652. Publisher Full Text\n\nMathis A, Perriere C, Zaccone G, et al.: In Fish defenses Volume 2: Pathogens, parasites and predators. eds G. Zaccone, C. Perriére, A. Mathis, & B. G. Kapoor. 2009; 2: 323–386 (Science Publishers). Reference Source\n\nMathis A, Smith RJF: Chemical alarm signals increase the survival time of fathead minnows (Pimephales promelas) during encounters with northern pike (Esox lucius). Behav Ecol. 1993; 4(3): 260–265. Publisher Full Text\n\nStabell OB, Lwin MS: Predator-induced phenotypic changes in crucian carp are caused by chemical signals from conspecifics. Environ Biol Fishes. 1997; 49(1): 139–144. Publisher Full Text\n\nWisenden BD, Smith RJF: The effect of physical condition and shoalmate familiarity on proliferation of alarm substance cells in the epidermis of fathead minnows. J Fish Biol. 1997; 50(4): 799–808. Publisher Full Text\n\nChivers DP, Smith RJF: Chemical alarm signalling in aquatic predator-prey systems: A review and prospectus (fait parti du numéro thématique sur l’écologie chimique/part of the special feature on chemical ecology). Ecoscience. 1998; 5(3): 338–352. Reference Source\n\nSmith RJF: Alarm signals in fishes. Rev Fish Biol Fish. 1992; 2(1): 33–63. Publisher Full Text\n\nMathis A, Chivers DP, Smith RJF, et al.: Chemical alarm signals: Predator deterrents or predator attractants? Am Nat. 1995; 145(6): 994–1005. Publisher Full Text\n\nChivers DP, Brown GE, Smith RJF, et al.: The evolution of chemical alarm signals: Attracting predators benefits alarm signal senders. Am Nat. 1996; 148(4): 649–659. Publisher Full Text\n\nCameron AM, Endean R: Epidermal secretions and the evolution of venom glands in fishes. Toxicon. 1973; 11(5): 401–410. PubMed Abstract\n\nAl-Hassan JM, Thompson M, Criddle RS, et al.: Composition of the proteinacous gel secretion from the skin of the Arabian Gulf catfish (Arius thallasinus). Mar Biol. 1982; 70(1): 27–33. Publisher Full Text\n\nChivers DP, Wisenden BD, Hindman CJ, et al.: Epidermal ‘alarm substance’ cells of fishes maintained by non-alarm functions: possible defence against pathogens, parasites and UVB radiation. Proc Biol Sci. 2007; 274(1625): 2611–2619. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamilton WD: The genetical evolution of social behaviour I. J Theor Biol. 1964; 7(1): 1–16. PubMed Abstract \| Publisher Full Text\n\nMaynard-Smith J: Group selection and kin selection. Nature. 1964; 201(4924): 1145–1147. Publisher Full Text\n\nTamura K, Ihara Y: Classes of communication and the conditions for their evolution. Theor Popul Biol. 2011; 79(4): 174–183. PubMed Abstract \| Publisher Full Text\n\nCharnov EL, Krebs JR: The evolution of alarm calls: Altruism or manipulation? Am Nat. 1975; 109(965): 102–107. Publisher Full Text\n\nSherman PW: Alarm calls of Belding ground squirrels to aerial predators: nepotism or self-preservation. Behav Ecol Sociobiol. 1985; 17(4): 313–323. Publisher Full Text\n\nda Silva KB, Mahan C, da Silva J, et al.: The trill of the chase: Eastern chipmunks call to warn kin. J Mammal. 2002; 83(2): 546–552. Publisher Full Text\n\nShelley EL, Blumstein DT: The evolution of vocal alarm communication in rodents. Behav Ecol. 2005; 16(1): 169–177. Publisher Full Text\n\nPollard KA: Making the most of alarm signals: the adaptive value of individual discrimination in an alarm context. Behav Ecol. 2010; 22(1): 93–100. Publisher Full Text\n\nThünken T, Meuthen D, Bakker TCM, et al.: Parental investment in relation to offspring quality in the biparental cichlid fish Pelvicachromis taeniatus. Anim Behav. 2010; 80(1): 69–74. Publisher Full Text\n\nLamboj A: Die Cichliden des westlichen Afrikas. (Birgit Schmettkamp Verlag) 2004. Reference Source\n\nThünken T, Bakker TC, Baldauf SA, et al.: Active inbreeding in a cichlid fish and its adaptive significance. Curr Biol. 2007; 17(3): 225–229. PubMed Abstract \| Publisher Full Text\n\nThünken T, Bakker TCM, Baldauf SA, et al.: Direct familiarity does not alter mating preference for sisters in male Pelvicachromis taeniatus (Cichlidae). Ethology. 2007; 113(11): 1107–1112. Publisher Full Text\n\nThünken T, Meuthen D, Bakker TC, et al.: A sex-specific trade-off between mating preferences for genetic compatibility and body size in a cichlid fish with mutual mate choice. Proc Biol Sci. 2012; 279(1740): 2959–2964. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThünken T, Waltschyk N, Bakker TC, et al.: Olfactory self-recognition in a cichlid fish. Anim Cogn. 2009; 12(5): 717–724. PubMed Abstract \| Publisher Full Text\n\nThünken T, Baldauf SA, Kullmann H, et al.: Size-related inbreeding preference and competitiveness in male Pelvicachromis taeniatus (Cichlidae). Behav Ecol. 2011; 22(2): 358–362. Publisher Full Text\n\nHesse S, Bakker TCM, Baldauf SA, et al.: Kin recognition by phenotype matching is family-rather than self-referential in juvenile cichlid fish. Anim Behav. 2012; 84(2): 451–457. Publisher Full Text\n\nMeuthen D, Baldauf SA, Bakker TCM, et al.: Substrate-treated water: a method to enhance fish activity in laboratory experiments. Aquat Biol. 2011; 13(1): 35–40. Publisher Full Text\n\nMcLennan DA: Male brook sticklebacks’ (Culaea inconstans) response to olfactory cues. Behaviour. 2004; 141(11): 1411–1424. Publisher Full Text\n\nMehlis M, Bakker TC, Frommen JG, et al.: Smells like sib spirit: kin recognition in three-spined sticklebacks (Gasterosteus aculeatus) is mediated by olfactory cues. Anim Cogn. 2008; 11(4): 643–650. PubMed Abstract \| Publisher Full Text\n\nBourdeau PE, Johansson F: Predator-induced morphological defences as by-products of prey behaviour: a review and prospectus. Oikos. 2012; 121(8): 1175–1190. Publisher Full Text\n\nKats LB, Dill LM: The scent of death: Chemosensory assessment of predation risk by prey animals. Ecoscience. 1998; 5(3): 361–394. Reference Source\n\nMateo JM: Recognition systems and biological organization: The perception component of social recognition. Ann Zool Fenn. 2004; 41(6): 729–745. Reference Source\n\nBrown GE, Adrian JC, Smyth E, et al.: Ostariophysan alarm pheromones: Laboratory and field tests of the functional significance of nitrogen oxides. J Chem Ecol. 2000; 26(1): 139–154. Publisher Full Text\n\nPfeiffer W, Riegelbauer G, Meier G, et al.: Effect of hypoxanthine-3(N)-oxide and hypoxanthine-1(N)-oxide on central nervous excitation of the black tetra Gymnocorymbus ternetzi (Characidae, Ostariophysi, Pisces) indicated by dorsal light response. J Chem Ecol. 1985; 11(4): 507–523. Publisher Full Text\n\nMathuru AS, Kibat C, Cheong WF, et al.: Chondroitin fragments are odorants that trigger fear behavior in fish. Curr Biol. 2012; 22(6): 538–544. PubMed Abstract \| Publisher Full Text\n\nRoh E, Mirza RS, Brown GE, et al.: Quality or quantity? The role of donor condition in the production of chemical alarm cues in juvenile convict cichlids. Behaviour. 2004; 141(10): 1235–1248. Publisher Full Text\n\nBrown GE, Bongiorno T, DiCapua DM, et al.: Effects of group size on the threat-sensitive response to varying concentrations of chemical alarm cues by juvenile convict cichlids. Can J Zool. 2006; 84(1): 1–8. Publisher Full Text\n\nBrown GE, Foam PE, Cowell HE, et al.: Production of chemical alarm cues in convict cichlids: the effects of diet, body condition and ontogeny. Ann Zool Fenn. 2004; 41(3): 487–499. Reference Source\n\nWard AJW, Hart PJB: The effects of kin and familiarity on interactions between fish. Fish and Fisheries. 2003; 4(4): 348–358. Publisher Full Text\n\nLangen K, Schwarzer J, Kullmann H, et al.: Microsatellite support for active inbreeding in a cichlid fish. PLoS One. 2011; 6(9): e24689. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "335", "date": "15 Oct 2012", "name": "Dustin Rubenstein", "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 premise of this study is based on a theoretical model (Pollard 2010, Behavioral Ecology) demonstrating that receivers might use signaler identity to infer signaler reliability.Such a framework may be appropriate when different signalers vary in their reliability, but in the case of alarm cues released from dead or wounded fish, the likelihood of dishonest alarm signals seem low. In other words, it is not clear why one would expect different responses to kin or non-kin alarm cues in these cichlid fish. Given that these cues must always honestly signal the presence of predator because of how they are released when an individual is wounded, fish would be expected to respond to any cue they find in the water. I therefore wonder if in this type of system that kin selection hypothesis is even appropriate to test. It is therefore not surprising to me that the authors find no difference in response to kin or non-kin alarm cues. I would have liked to have seen more discussion about the appropriateness of testing this model in this type of signal-receiver system.", "responses": [ { "c_id": "838", "date": "27 May 2014", "name": "Denis Meuthen", "role": "Author Response", "response": "Rubenstein criticized the reference to the theoretical model by Pollard 2010. We realized that the reference to the Pollard paper was a bit misleading and may have caused confusion. In our revision, we replaced this aspect with a more detailed explanation of why we expect kin selection to play a role in the evolution of alarm cues. We now include other studies previously highlighting the potential role of kin selection in the evolution of alarm signals, including fish alarm cues. A particularly interesting example on plants shows that sagebrush induces strong anti-herbivore protection mechanisms when exposed to the volatile (alarm) cues of wounded close relative plants. In contrast, the anti-herbivore response was weaker when they were exposed to volatiles from distantly related plants.Second, Rubenstein raised concerns that due to physiological restrictions of alarm cues in fish, dishonest alarm signals are unlikely. We fully agree with the referee. However, it was not our intent to convey that our study is based on the premise of the dishonest or honest alarm signals which are referenced by Pollard 2010. Instead, we focus on the question whether kin are able to respond more quickly towards a predator by being more sensitive towards kin alarm cues. For this purpose, in our revision we present an argument outlining that the recognition of kin alarm cues could be based on an indirect coupling of concurrently present kin-specific cues and alarm cues. Accordingly, fish could be able to discriminate between kin- and non-kin alarm cues based on learned kin recognition. This would not be surprising as kin recognition has been shown to be present in fishes and injuries caused by predators release numerous substances contained in very different areas of the prey body." } ] }, { "id": "336", "date": "02 Nov 2012", "name": "Maurice W Sabelis", "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\nMeuthen et al. describe the results of experiments to test whether the cichlid fish Pelvicachromis taeniatus discriminates between alarm cues produced by kin and non-kin upon being injured by a predator.They manipulated alarm cues (as a proxy of predation risk) in three treatments by exposing individual fish to one of the following: (1) Alarm cues derived from kin (from familiar and unfamiliar siblings); (2) Alarm cues derived from unfamiliar, unrelated conspecifics (non-kin); (3) Distilled water to control for disturbance effects upon introduction. They showed that activity of fish receiving alarm cues from treatment 1 and 2 was reduced compared to the control (treatment 3), but familiarity of siblings (treatment 1) and kinship (treatment 1 vs 2) had no effect on activity. Meuthen et al. conclude that P. taeniatus fish do not discriminate between alarm cues from kin and non-kin. Finally, they argue that kin selection can only play a role if this fish usually lives in groups of kin and that alarm cues are more likely to have evolved as a by-product of wound-healing or immune responses to pathogens colonizing cells wounded by predators.While their experiments are clearly described and carefully designed, we question whether the question in the title of the paper by Meuthen et al. is well posed: Evolution of alarm cues: a role for kin selection? First, alarm cues represent public information and we see no reason why kin and non-kin should respond differently to these cues (except in the very special case where predation risk differs between kin and non-kin). Hence, we disagree where the authors state: “Higher sensitivity to kin-alarm cues may result in an improved response to predation, and thus higher survival of individuals related to the sender which in turn may increase the indirect fitness of the sender”. Why would a higher sensitivity to kin alarm be expected in the first place? Second, the experiments by Meuthen et al. were designed such that they exclude the possibility for wounded fish to decide on what and how much to send depending on whether it is surrounded by kin or non-kin. This presupposes that the wounded fish has some form of control over the amount of alarm cues that are released. In our view this extent of this sender control is an essential target for kin selection. Thus, to answer the question posed in the title of the paper the experiments should have been designed so as to allow the potential sender to perceive who are the potential receivers and so as to measure the amount of alarm cues released depending on their environment.Side remarks:(1) In the Materials and Methods section the authors state: “(experimental stimuli) … were introduced at the point of 1h 15min and then fish behavior was recorded for another hour. Thereafter, all experimental subjects were sized accurate to the nearest millimeter and weighed accurate to one milligram on an electrical precision scale (LC 2215, Sartorius, Germany)”. However, the results of these measurements are not described in the paper.(2) In the Results section the analysis focuses on activity as the difference in distance covered by the fish before and after application of the treatment. These values may differ (as they do between treatments 1 and 2 vs 3) due to differences before treatment, but the authors seem to assume that any difference in this activity parameter is due to differences after application of the treatment. We strongly suggest the authors to provide a statistical test to show that the distances before treatment were not significantly different among the three groups.", "responses": [ { "c_id": "837", "date": "27 May 2014", "name": "Denis Meuthen", "role": "Author Response", "response": "Sabelis et al. critically remarked \"whether the question in the title of the paper is well posed\". To our opinion the evolution of alarm cues is puzzling as their production is a costly process (individuals with higher condition produce more alarm cues). If individuals benefit from alarm cues while not contributing to the alarm system by producing own alarm cues, they would increase in number and ultimately eliminate this alarm signaling system from evolution. Kin-biased perception could solve this problem. In accordance with this hypothesis a recent paper shows that discrimination between kin and non-kin volatile chemical (alarm) cues occurs in plants (Karban et al. 2013). Moreover, several studies suggest that familiarity with alarm signals (O'Connell-Rodwell et al. 2007) or chemical cues (Coopersmith & Leon 1984, Brown & Smith 1994) leads to improved responses. Along the hypotheses of other researchers concerning the role of kin selection in the evolution of mammalian alarm calls (Sherman 1977, Charnov & Krebs 1975, Sherman 1985, da Silva et al. 2002) and alarm cues of fishes (Smith 1992), we therefore expect that individual fish should respond differently to alarm cues derived from kin and non-kin.Furthermore, Sabelis et al. suggest that \"wounded fish have some form of control over the amount of alarm cues that are released, ultimately providing an essential target for kin selection\". If wounded fish had the control, this would be indeed a possibility. However, it is usually assumed that alarm cues are passively released. Alarm cues are putatively located in enclosed subepidermal club cells without external ducts, thus providing little opportunity for short-term variation in the released alarm cue amount. Accordingly, the only way for alarm cues to be released is the destruction of these cells by predators. Therefore, we focused on the receiver part, i.e. whether individuals are differentially sensitive towards kin and non kin alarm cues. One possible control by the sender might be differential long-term investment in club cell quantity depending on whether the sender is surrounded by relatives. In the present study all fish grew up in kin groups prior to trials, therefore, we exclude the possibility of differential investment based on the long-term sender environment.Following Sabelis et al. side remarks we included additional analyses regarding fish sizes and activities before the treatment. These analyses showed that neither fish size nor weight nor prestimulus activity differed significantly among the three treatment groups." } ] } ]	1	https://f1000research.com/articles/1-27
https://f1000research.com/articles/3-117/v1	27 May 14	{ "type": "Case Report", "title": "Positioning a proned patient with cauda equina syndrome who presents at 15 weeks gestation: a case report", "authors": [ "Elizabeth Speirs", "Matthew Wiles", "Andrew Bacon", "Stephen Radley", "Matthew Wiles", "Andrew Bacon", "Stephen Radley" ], "abstract": "Cauda equina syndrome is a neurosurgical emergency that requires prompt intervention to prevent irreversible spinal cord paralysis. This article describes how we managed a case of an obese pregnant patient who was placed in the prone position for surgery. We discuss the evidence behind the management options and choice of operating tables available.", "keywords": [ "We present a case of Cauda equina syndrome in a 24 year old woman at 15 weeks gestation. She had no other medical problems to note and had previously had two uncomplicated pregnancies. Her elevated body mass index (BMI) of 36 kg/m-2 did provide a potential difficulty in managing this case." ], "content": "Case\n\nWe present a case of Cauda equina syndrome in a 24 year old woman at 15 weeks gestation. She had no other medical problems to note and had previously had two uncomplicated pregnancies. Her elevated body mass index (BMI) of 36 kg/m-2 did provide a potential difficulty in managing this case.\n\nThe patient was referred to our tertiary neurosurgical unit with a 24 hour history of severe lower back pain, altered perianal sensation and right lower limb weakness. An urgent MRI scan showed a large L5/S1 disc prolapse (Figure 1). A frank discussion of the risks of medical versus surgical management was undertaken, and immediate surgery was chosen by the patient.\n\nThe decompression was performed with our patient proned under general anaesthesia, which was maintained using a combination of sevoflurane and remifentanil. Due to the body habitus of the patient (BMI 36 kg/m-2), it was felt that undertaking surgery in the lateral position would have been technically difficult, with an increased operative time and a greater risk of bleeding. We chose to prone the patient on to a Jackson table (OSI, Union City, CA) with supports on the sternum and laterally on the iliac spines. Absolute care was taken to ensure that there was no abdominal compression once positioned.\n\nThe operation was performed uneventfully and lasted 90 minutes. Cardiovascular stability was maintained throughout. The patient’s neurology had fully recovered by the first postoperative day and a fetal ultrasound that day showed a viable fetus. Unfortunately this was later found to be a twin pregnancy, one of whom was spontaneously aborted. The second twin was delivered healthy at 39 weeks gestation.\n\n\nDiscussion\n\nThere is little literature currently available to guide positioning, especially discussing proning techniques in pregnant patients for spinal surgery. Discectomies in gestational patients can be performed in both the prone and lateral position1. A proned patient allows better surgical access, but the lateral position may make it easier to ensure that there is no abdominal compression to compromise uterine blood flow. Even though the majority of operating tables designed for proning allow the abdomen to hang free in non-pregnant patients, they are not designed to fit the larger pregnant abdomen. We could find no literature comparing the use of different operating tables and the abdominal space that they offer a pregnant abdomen.\n\nUterine blood flow during prone positioning has been examined by Nakai et al.2. They took 23 healthy women at 34 weeks gestation and laid them in a supine, right lateral, left lateral and prone position. The proning table had a hole in it for the pregnant abdomen so they could see that no compression occurred. They found that prone positioning actually provided optimal relief of umbilical artery compression as measured by the umbilical artery systolic/diastolic ratio, and that this is superior to that of both the right and left lateral positions. The effect of the prone position on uterine blood flow in an anaesthetised patient has not been determined however.\n\nThe few small case series and cases that we found describing spinal surgery in the prone position for gestational women have demonstrated good fetal outcomes. One series3 presents the course of three women of 16 to 20 weeks gestation who required discectomies for lumbar disc herniation during pregnancy. They all received epidural anaesthesia and then self positioned themselves prone on a Relton-Hall laminectomy frame (IS, Dorval, Quebec). All babies were delivered without complications at full term. The authors chose the Relton-Hall frame as it has pressure points on the anterior superior iliac spines and the chest, which they felt allowed most freedom of the abdomen and uncompressed Inferior vena cava blood flow. In addition, they concluded that letting the women position themselves prior to receiving an anaesthetic was an extra safety measure. They believed that if they could ensure that the women were comfortable once proned, they were satisfied that there was no undue pressure on the abdomen and uterus. Even at 30 weeks gestation, the technique of regional anaesthesia and patient self-positioning has been seen to have no adverse foetal effects4. In this case they also undertook foetal monitoring throughout surgery, and noted no foetal compromise intra-operatively.\n\nCervical and thoracic decompressions for epidural haematomas are similar surgical emergencies that present the need to prone patients under general anaesthesia. A case series of six patients reports all the babies being born healthily after spinal decompressions5. These patients were placed on a Wilson frame (ZA, Jiangsu, China). Three procedures were undertaken at 20, 24 and 34 weeks, with spontaneous deliveries occurring at term. Three other women of 35, 38 and 41 weeks gestation underwent caesarean sections prior to decompression with no adverse foetal outcomes. The recommendations from this series are that neurosurgical intervention to avoid permanent neurological damage is safe and that surgery should not be delayed in obstetric patients.\n\nThere is only one published guideline that we could find on positioning and spinal surgery in gestational women, produced by a centre in Korea1. They followed a case series of ten women from their institution, 6 of who had lumbar disc herniation, the others infection or tumour. All patients in the first trimester were placed in a prone position, five delivering healthy full term babies whilst one patient had a therapeutic abortion for early radiation exposure. Of the later term patients, two were put in right, and two in left lateral positions. One baby was born healthily at full term, one by elective caesarean section at 34 weeks gestation, one pre-term at 33 weeks spontaneously and one therapeutic abortion was performed, also due to radiation exposure. Their guidelines produced from this review suggest epidural anaesthesia for most procedures, but for longer operating duration general anaesthesia may be preferable. It also recommends that during the first and early second trimester a prone position for surgery is safe, but a left lateral position is preferable for the latter part of the second trimester and third trimester. They conclude that from their review and own experiences peripartum, neurosurgical procedures can be safely performed in most pregnancies.\n\n\nConclusion\n\nIn summary, there is very limited evidence to guide the optimal management of the pregnant patient who requires emergency spinal surgery. However, safe outcomes for both mother and fetus undergoing general anaesthesia for varying pathologies during the first trimester have been seen in many case series6. These women were all supine, but there appears to be no reason to delay general anaesthesia during the first trimester. For pregnancies of 34 weeks gestation and over, a caesarean section can be performed safely prior to, or, as a combined procedure with discectomies7.\n\nAnother very important aspect of case management involves a multi disciplinary team approach peri-operatively. A fetal ultra sound should be performed before any intervention is undertaken and fetal heart monitoring intra-operatively should be considered. Alterations in fetal heart rate can alert the anaesthetist as early as possible to aortocaval compression and cardiovascular insufficiency due to poor positioning of the mother. Close obstetric follow up is essential in any non-obstetric surgery.\n\nFrom all the literature that we have looked at, there appears to be no reason to avoid the prone position for surgery. Nor is there evidence to suggest that regional or general anaesthesia offers significant benefits to either mother or foetus. There is also no particular operating table that we can recommend, but knowledge of your own equipment is vital. We believe that the key aspects of the management of such cases is to take meticulous care during positioning to ensure that the abdomen is free, no matter what position or operating table is chosen.\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\nES wrote up the case report and first draft and MW was the main reviewer. AB and SR provided expert input in the case and reviewed the case report.\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\nHan IH, Kuh SU, Kim JH, et al.: Clinical approach and surgical strategy for spinal diseases in pregnant women: a report of ten cases. Spine. 2008; 33(17): E614–619. PubMed Abstract \| Publisher Full Text\n\nNakai Y, Mine M, Nishio J, et al.: Effects of maternal prone position on the umbilical arterial flow. Acta Obstet Gynecol Scand. 1998; 77(10): 967–969. PubMed Abstract \| Publisher Full Text\n\nBrown MD, Levi AD: Surgery for lumbar disc herniation during pregnancy. Spine. 2001; 26(4): 440–443. PubMed Abstract\n\nKim HS, Kim SW, Lee SM, et al.: Endoscopic discectomy for the cauda equina syndrome during third trimester of pregnancy. J Korean Neurosurg Soc. 2007; 42(5): 419–420. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJea A, Moza K, Levi AD, et al.: Spontaneous spinal epidural haematoma during pregnancy: Case report and literature review. Neurosurgery. 2005; 56(5): E1156. PubMed Abstract\n\nCohen-Kerem R, Railton C, Oren D, et al.: Pregnancy outcome following non-obstetric surgical intervention. Am J Surg. 2005; 190(3): 467–473. PubMed Abstract \| Publisher Full Text\n\nBrown MD, Brookfield KF: Lumbar disc excision and cesarean delivery during the same anesthesia. A case report. J Bone Joint Surg Am. 2004; 86-A(9): 2030–2203. PubMed Abstract" }	[ { "id": "4905", "date": "02 Jun 2014", "name": "Bernard Wittels", "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\nThese authors present an unusual case of an obese, pregnant woman at 13 weeks gestational age who presented with perineal hypesthesia and partial loss of motor function in her right lower leg. An MRI showed an L5-S1 herniated nucleus pulposus that required emergency decompression. The authors discuss their anesthetic management of the patient and review the current literature.Title: The title should be re-worded because the patient did not have cauda equina syndrome (which usually includes partial or full loss of control of bowel and bladder sphincters, and may include painful paresthesias), and the word “prone” should only be used as an adjective. A more appropriate title would be: Anesthesia for emergent lumbosacral discectomy and nerve decompression in an obese parturient at 13 weeks gestation.Case report:Again, the term cauda equina should be removed. The list of potential difficulties in administering anesthesia to an obese, preterm parturient needs to be elaborated in detail. There is no mention of consenting the patient for the MRI, including the risks of MRI to the developing fetus. The patient's sensorimotor radiculopathies represent a neurosurgical emergency; medical therapy in this setting does not conform to the standard of care and should not have been offered as a valid alternative therapy. To conform with their discussion points, the authors should have obtained an obstetric consultative evaluation. The absence of preoperative, intraoperative, and postoperative obstetric consultations should be explained. The failure to detect twin gestations at 13 weeks of age, despite having an MRI, needs to be discussed. It appears that epidural and/or spinal anesthesia were not considered or offered as an alternative to general anesthesia, nor were these two anesthesia techniques compared for their relative risks and benefits to the mother and her fetus. The authors should mention that spinal anesthesia produces decreased fetal drug exposure and decreased maternal blood loss, compared to general anesthesia. What was the patient's airway evaluation? Many obese parturients have difficult airways that can pose life-threatening problems with hypoxemia and failure to intubate after induction of anesthesia. How was the anesthetic induction managed? What drugs were used? What happened to her oxygenation, ventilation, and hemodynamics on induction of anesthesia? Why were sevoflurane and remifentanil chosen for maintenance of anesthesia? What advantages or disadvantages does this regimen have over other choices? In an obese patient, the prone position would seem to be the only reasonable choice by the neurosurgeon. Abdominal compression is just one consideration in the safe positioning of an obese, pregnant patient prone on the operating table. What are the other concerns? Could the fetal heart rate be monitored externally? If so, then why wasn't fetal heart monitoring employed intraoperatively? What would be the expected fetal heart rate pattern? Exactly when was the twin gestation diagnosis made? Why wasn't that diagnosis made earlier? Could the radiation from the MRI, or the effects of surgery, anesthetic drugs, or the prone position have any potential effect on halting the growth and development of the unborn twin? Discussion: It would be useful for the authors to distinguish known safety concerns in this setting from factors that are unknown and untested. This section should be shortened. The Nakai reference should be dropped because it was done in awake patients near term gestation. No comparative study of OR tables for the prone position of pregnant patients has been done. It is inconsistent to review a study that used intraoperative fetal monitoring without commenting on why it was not used on your study case patient. None of the referenced studies involved preterm gestations or obese parturients. ASA guidelines are worth referencing, but not one from Korea based on a single study of ten patients.Conclusion: What did you do differently throughout the perioperative period to manage these problems that were unique to your case? Where did you succeed? After reviewing the literature, how do you see your case enhancing our knowledge of neuroanesthesia during pregnancy? Since there are no studies comparing regional versus general anesthesia in this setting, no conclusion can be made about their relative benefits and risks. Abdominal compression alone is not the only major concern with the prone position. What are other concerns and how are these assessed and managed? What would you do differently or better in the future to provide the safest level of care? Can you suggest further research that is needed?", "responses": [] }, { "id": "4956", "date": "02 Jun 2014", "name": "John Svigos", "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\nWhilst I find it to be an informative and current paper with a good appraisal of the available literature, intuitively and grammatically I believe the use of \"proning\" etc is incorrect as prone is an adjective and must be used in this manner as the \"alternative\" use detracts from the paper.", "responses": [] }, { "id": "4903", "date": "11 Jun 2014", "name": "Michael Paech", "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 submission is a case report of a 15 week pregnant woman who needs urgent lumbar discectomy for cord compression. This requires the use of the prone position (PP) on a Jackson table.The use of a so-called Jackson Table is sensible. It allows for the abdomen to hang freely between the pelvic and chest support. It thus reduces the epidural venous pressure that may cause troublesome bleeding for the surgeon. The Wilson frame (as mentioned in the Discussion) will cause more pressure on the abdomen.The main message is the discussion of the effects of the PP on uteroplacental blood flow and fetal outcome. However, in week 15 the effects on the uteroplacental blood flow may not be significant and the other problems associated with the prone position should be mentioned.A number of other case reports and case series are presented that indicate the PP may be safe during pregnancy. However, the lack of reference to the anaesthetic literature in the references suggests that the authors may not be aware use of the PP during GA is already reasonably well described during pregnancy. More specific comments:The word ‘proning’ (and term ‘proned patient’ ) are incorrect in our opinion – use as a verb except in gaming vernacular. The terms ‘prone’ or ‘patient in the prone position’ should be substituted throughout. There is inadequate discussion of the anaesthetic issues. Although “safe outcomes” occur after GA in the first trimester, the clear consensus in the anaesthesia community is to avoid or delay GA until the second trimester if possible and to use regional anaesthesia when possible. Conclusion:L2-3: What is meant by “varying pathologies”? Do the authors refer to spinal pathology or any pathology that requires general anaesthesia during first trimester?L7-8: Fetal heart rate monitoring was not adequately discussed. During surgery is not useful until after fetal viability and is difficult to interpret at early gestations. It may give some indication of fetal compromise that prompts change in anaesthesia, but it is only of true value if provision has been made to perform emergency delivery, which does not apply in this case.", "responses": [] } ]	1	https://f1000research.com/articles/3-117
https://f1000research.com/articles/3-59/v1	18 Feb 14	{ "type": "Research Article", "title": "Spatial response of the globally-endangered Sokoke Pipit (Anthus sokokensis van Someren, 1921) to habitat modification in an Eastern Arc Coastal Forest", "authors": [ "Nickson Erick Otieno", "David Ngala", "Alex Mwalimu", "David Ngala", "Alex Mwalimu" ], "abstract": "The Arabuko-Sokoke forest is the largest relic of a formerly larger contiguous East African coastal forest. It forms part of the Eastern Arc Forest system which is a global biodiversity hotspot with considerable species endemism. As one of the most important bird habitats in Africa, it hosts nine globally-threatened and four regionally threatened species. Despite such conservation significance, the forest is undergoing rapid modification and habitat loss mainly from anthropogenic pressures, with negative impacts on sensitive species such as the Sokoke Pipit (Anthus sokokensis). This study examined impacts of change in habitat quality on the species’ population and spatial occurrence within three blocks of Brachystegia woodland in the forest. Over a three week period, six 1-km transects were used to estimate the species’ population in relation to major habitat quality variables. Sokoke Pipits occurred at an overall mean density of 0.72±0.15 birds/ha with an estimated population of 5,544 in the Brachystegia woodland. Tree logging intensity was the key cause of the degradation of the Sokoke Pipit’s critical habitat, which affected its density (R2 = 0.663, ß = -0.814, p = 0.048). The species also preferred sites covered with deep floor litter (R2 = 0.769, ß = 0.877, p = 0.021) even in areas with low tree canopy height, but showed no clumped distribution (χ2(2, 0.05) = 2.061). Sites with intensive elephant activity, which leads to tree felling and clearing of the understorey, had low Sokoke Pipit densities. We conclude that although human-driven tree removal is a major driver of degradation of the Sokoke Pipit’s critical habitat, elephant activity may be an important additional factor in this process. Long term conservation strategies for the species will require stricter control of logging but management of the population and dispersal of elephants across the forest, especially in Brachystegia woodland, may also be helpful.", "keywords": [ "Tropical forests constitute the most important habitats for biodiversity because despite covering less than 7% of the global land surface", "they host at least half of all terrestrial species on earth1. However", "these habitats also face the greatest threat from human exploitation", "destruction or modification with an estimated loss rate of c. 10% every decade especially in areas without formal protection1", "2. Birds are among the most affected by forest destruction and habitat loss", "particularly forest-dependent species3", "4 which may respond to such perturbations in such spatially distinct patterns as to make them suitable for monitoring the quality of the forest habitat and its suitability for other taxa5." ], "content": "Introduction\n\nTropical forests constitute the most important habitats for biodiversity because despite covering less than 7% of the global land surface, they host at least half of all terrestrial species on earth1. However, these habitats also face the greatest threat from human exploitation, destruction or modification with an estimated loss rate of c. 10% every decade especially in areas without formal protection1,2. Birds are among the most affected by forest destruction and habitat loss, particularly forest-dependent species3,4 which may respond to such perturbations in such spatially distinct patterns as to make them suitable for monitoring the quality of the forest habitat and its suitability for other taxa5.\n\nSokoke Pipit6 is a forest-floor insectivore of the East African coastal forests of Kenya and Tanzania7–9. This globally-endangered species10 is generally restricted to the woodland habitat dominated by Brachystegia tree species (Leguminoceae)9,11 where it feeds on arthropods on the ground or in the understorey9,12–14. It mainly occurs in parts of the woodland with deep floor litter and a fairly closed tree canopy9.\n\nThe species has been more frequently encountered in the coastal forests of Kenya than those of Tanzania, with the most common sites being in the Arabuko-Sokoke forest, hereafter ASF and the Dakatcha Woodland. The main threat to the species is the modification and reduction of its suitable habitat12,14, especially the removal of Brachystegia trees, a process to which it is very sensitive9.\n\nThis study aimed to assess the Sokoke Pipit’s response to change and modification of its habitat by comparing its estimated densities across three zones of its Brachystegia woodland stronghold in ASF. Although there have been extensive previous studies on the forest’s biodiversity in general15,16; on other forest-dependent bird species17,19,20 and one on forest disturbance11 no study has been conducted to directly investigate the link between Sokoke Pipit population or distribution and modification of its habitat. Musila et al. examined14 the species’ general habitat requirements within Brachystegia woodland, but did not specifically examine its demographic and spatial response to change in structural habitat quality. Our study thus aimed at filling these gaps as well as providing updates on the species’ current population estimate over the past decade since the study by Musila et al.14 in 2001. In that earlier study species density was estimated at 2.8 ha-1 and 0.7 ha-1, in the undisturbed and disturbed areas of the Brachystegia forest respectively, and overall population at 13,000 for the whole forest.\n\n\nMaterials and methods\n\nThe ASF is located between 39°40′E–39°50′E longitude and 3°10′S–3°30′S latitude, within the Malindi and Kilifi Districts along Kenya’s north coast (18 km south of Malindi) see Figure 1. Its altitude ranges from 60 to 200 m above sea level20, and mean annual rainfall from 600 mm in the northwest to 1100 mm in the northeast, with the rainy season falling between late March and May, the short rains occurring from November to December and dry season from June to October and December to February11. Mean monthly temperatures range from 26 to 31°C. The forest is one of the few remaining indigenous forests in Kenya, and one of the largest fragments of an earlier, much larger coastal forest that once covered much of the East African coast21. It covers 41,600 ha18 including 4,300 ha which is formally protected as a nature reserve9,20.\n\nThe figure shows the map of Arabuko-Sokoke forest indicating the main blocks in Brachystegia woodland where surveys were conducted (Jilore, Narasha and Kararacha) and the six transects used (two in each block). Numbers 1, 2, 3…. are transects numbers in the blocks (Map adapted from Davis, 2005)20.\n\nThe ASF constitutes one of the Eastern Arc Coastal biodiversity hot spots22 and is one of the most significant Important Bird Areas in Kenya based on BirdLife International protocols9,15. It hosts at least 230 bird species including five globally-endangered species (Sokoke Pipit Anthus sokokensis, Spotted Ground Thrush Zoothera guttata, Sokoke Scops Owl Otus ireneae, Clarke’s Weaver Ploceus golandi and Amani Sunbird Anthreptes pallidigaster); four near-threatened species (East Coast Akalat Sheppardia gunningi, Plain-backed Sunbird Anthreptes reichenowi, Fischer’s Turaco Tauraco fischeri and Southern-banded Snake Eagle Circaetus fasciolatus); and eight regionally-vulnerable species9,10,18. Among these are five species endemic to the East African Coastal Forests Endemic Bird Area: Sokoke Pipit, Clarke’s Weaver, Amani Sunbird, Fischer’s Turaco and Sokoke Scops Owl9,14,16. These features make ASF the second most important forest for bird conservation in mainland Africa4. In addition there are three endangered mammals including the African elephant (Loxodonta africana), an endangered amphibian Mertensophryne micrannotis and the rare lizard Gastropholis prasina as well as at least 50 globally rare plant taxa9.\n\nThe ASF consists of three main forest plant community zones: Brachystegia woodland which runs in a central strip, a relatively open habitat dominated by Brachystegia spiciformis trees growing in low density mainly on whitish-leached sandy soils covers some 7,700 ha of largely open understorey with little or no undergrowth24; mixed forest with a diversity of relatively densely populated, tall and undifferentiated trees covering an area of about 7,000 ha; and Cyanometra forest with a variegated thicket, which covers about 23,500 ha, grows in the western part of the forest on red Magarini sands and is dominated by Cyanometra webberi, Manilkara sulcata, Oldfieldia somalensis and Brachystegia huillensis9,11,25. The rest consists of plantation forest and open gaps (Figure 1).\n\nThe forest is surrounded by small-scale agricultural land and settlement by a growing population of adjacent communities whose relatively low income levels are partly responsible for their increasing dependence on forest products for many of their needs9,26,27. According to current strategic forest management plan estimates, there are almost 60 villages scattered around the forest that utilize natural products directly derived from the forest9,18,25.\n\nThe main forms of human activities that impact on the forest habitat include illegal logging, honey harvesting, game snaring, cattle grazing and the creation of numerous tracks used by tree poachers into and out of the forest25. Further impact on the habitat is caused by the foraging behaviour of the resident population of African elephants, which has increased from an estimated mean value of 141 individuals in 199628,29 to between 180–227 in the period from 2002–200629,30. We were not able to obtain any official figures since 2006.\n\nThis study was conducted in the forest woodland zone dominated by Brachystegia spiciformis trees, to examine the response of the Sokoke Pipit population to human-induced habitat modification, particularly the removal of trees and other habitat degradation effects. The response to such variables was assessed in terms of the species’ density, encounter rates and distribution. Accordingly, we expected the species’ density and distribution to reflect corresponding spatial patterns in logging intensity.\n\n\nSampling strategy\n\nThe survey was carried out over 28 days between November 2011 and February 2012 within three blocks in the Brachystegia woodland stratified as follows: the main forest reserve block in the north-east, centered around Narasha (generally regarded as the most highly disturbed area from earlier intensive lumbering which continued until the early 1980s); the southern block of regenerating forest (a reserve regarded as less disturbed) in the Kararacha area; and the smaller strip on the outer north-western part of the forest around the Jilore village, which is considered more disturbed than Kararacha but slightly less so than Narasha (Figure 1). This classification is based on the methodology of Oyugi et al.11. Two 1-km transects were laid randomly in each block. Randomization was achieved by selecting the third track that branched to the left of the main forest track each time31,32. When such a track was too short to cover one whole kilometer of forest as was in the case in the Jilore zone, a track was selected to run parallel to the main forest track but maintaining at least 250 m from the main track and the forest edge. In addition, bird surveys were conducted by starting from a different end of the transect each successive time31. Sampling independence for bird detection was ensured by maintaining at least 1 km from neighboring transects. Bird surveys, vegetation sampling and habitat assessment for tree logging intensity were assessed on separate days.\n\nSokoke Pipit survey was the main objective of the study but we also recorded other birds encountered along the transects. The survey was conducted using Distance protocol, as described by Buckland et al.33 starting from 6.00 am to 9.00 am along the randomly selected 1-km transects in each forest block. Transect widths were fixed at 60 meters and birds were counted by moving slowly and recording all sightings and calls5,31. Surveyors worked in pairs, one observing with a pair of Bushnell XLT binoculars with 8×32 magnification and the other recording any encounters as they walked along the transect. Only positively identified Sokoke Pipit individuals or clusters were recorded. Perpendicular distance of each encounter from the transect centre was also determined, using a Nikon NKU 8371 rangefinder and recorded [see Buckland et al.33 and Fewster et al.34]. To reduce biases associated with double counting, birds flying from behind the surveyors were ignored and a distance of no less than 1 km was maintained between transects31. For clusters of birds, the perpendicular distance measured was to the centre of the point where the individual cluster was originally detected31,33.\n\nVegetation parameters were assessed within ten 10 × 10 m quadrats along the same transects used for birds. The quadrats were established on alternate sides of the transects at 100-m intervals and within them estimates of percent canopy height, using a Nikon NKU 8371 rangefinder, and canopy cover from three different points along a diagonal line down the quadrat (each corner and the centre) were scored. Canopy cover percent was scored within three defined ranges as < 33%; 34–66%; or > 66%). Live woody stems were also counted in each quadrat to gauge the understorey woody vegetation density. These were scored in three size classes of small (10–30 cm); medium sized (31–60); cm and large (above 60 cm) measured using a standard tape measure at breast height. In addition, logging intensity was assessed in each of the quadrats by counting all cut stems of trees of similar diameter size using the categories above11.\n\nIn each of the quadrats along transects used for vegetation sampling, forest floor litter depth was assessed at three points along a diagonal running from one corner to another through the quadrat centre32. The depth of litter was determined using a straight, stiff thin metallic rod driven vertically and gently downward until it touched the firm forest floor beneath the litter, and then read off against a standard 30 cm ruler. Litter cover was assessed by dividing the 10 × 10 m quadrats into 25 smaller grids of 2 × 2 m quadrats by use of a standard meter rule and tape measure then ascertaining the percentage category in each 2 × 2 square before averaging the total out of 25.\n\n\nData analyses\n\nBecause of the relatively small number of replicates in the study (two transect runs for birds and one set of habitat variable samples) preliminary data exploration showed departure from normal distribution. As such, all count data such as for live stems and cut tree stumps were transformed by logarithm and ratio or scale data such as by arc-sine before analyses proceeded31,35. Sokoke Pipit densities were determined per hectare using DISTANCE v 6 software33 while their encounter rates were also calculated from the relationship RE = n/Lt where RE = encounter rate; n = mean abundance of Sokoke Pipits along the transect; and Lt = total length of transect in kilometers. Due to high variance in detection of Sokoke Pipit in the Jilore block compared to Kararacha and Narasha (Table 1), which was likely a result of differences in understorey structural characteristics, Multiple Covariate Distance Sampling (MCDS) was preferable to Conventional Distance Sampling in estimating Sokoke Pipit density even with the relatively small sample size as the mean cluster sizes were quite constant at two individuals per sighting36,37. We selected the cosine adjusted half-normal detection functional model with the lowest value based on Akaike Information Criterion in the density estimations37. Species richness for all birds was evaluated as the total cumulative number of different species recorded in each transect during all the bird sampling sessions. Bird diversity was worked out using the reciprocal of Simpson’s index of the form: 1/S = 1/[(Σn(n-1)/N(N-1)] where S = Simpson’s Index, n = the total number of organisms of a particular species and N = the total number of organisms of all species. Simpson’s index of diversity was chosen as it is suitably robust for non-numerous replicate sampling such as was the case in the study32,35.\n\nThe mean number of live stems and tree stumps/cut stems were derived from all stems counted in the three size classes in all quadrats in transects and transformed into densities per hectare. Percent canopy cover scores were coded such that open canopy, moderately open canopy and closed canopy scored 1, 2 and 3, respectively. These were then transformed to ratios scaled with ‘3’ as the maximum. Canopy height, floor litter cover and litter depth measurements were averaged from all quadrats in all transects.\n\nDue to high preliminary-test covariance amongst the various size classes of live tree stems and tree stump counts, the size classes were pooled together into ‘total live stems’ and ‘total stems cut’ for subsequent analyses. For habitat variables that showed particularly strong correlations to bird variables, simple linear regression was performed to test the actual correlations and relative strengths of predictability. Differences of means of the key habitat (independent) variables were compared on the spatial scale by one-way Analysis of Variance (ANOVA) using the forest blocks as the categorical treatment effects on the bird (response) variables. The relationships between the independent and response variables (ANOVA and regressions) were analyzed in SPSS version 18.\n\n\nResults\n\nIn all surveys, a total of 308 birds were encountered, distributed across 55 species belonging to 25 families. This included three of the globally-endangered species: Sokoke Pipit, Clarke’s Weaver and Amani Sunbird; two globally near-threatened species: East Coast Akalat and Plain-backed Sunbird; and one regionally-vulnerable species: Little Yellow Flycatcher Erythrocercus holochlorus (Supplementary Table 1). There were 17 encounters of Sokoke Pipit with an overall abundance of 30 individuals. The Sokoke Pipit occurred at a mean overall density of 0.72 birds/ha across the blocks surveyed, with a projected overall population estimated at 5,544 individuals (Table 1). The density was higher in the moderate to high disturbance Brachystegia forest zone represented by Jilore and Kararacha blocks (0.89 birds ha-1) compared to the more disturbed Narasha block (0.71 birds ha-1). Nevertheless, there was no significant evidence of clumped distribution of the species across the blocks (χ2(2, 0.05) = 2.061).\n\nAIC = Akaike Information Criterion with right-truncated distances and cosine adjustment function.\n\nSimilarly, the species had the highest encounter rate in Jilore (4 km-1) while Kararacha had 2.0 birds km-1 and Narasha 1.5 birds km-1. The mean cluster size observed for Sokoke Pipit was 2 birds. For all birds, Kararacha had the highest species diversity (1/S = 0.69) followed by Narasha (1/S = 0.721) then the Jilore area (1/S = 0.724), S being the reciprocal of Simpson’s diversity index. Jilore was the most bird species-rich (38 species) followed by Narasha (35 species) and then Kararacha (34 species).\n\nFloor litter was deepest in the Kararacha block (2.52±0.83 cm) followed by the Jilore block (2.21±0.73 cm) and Narasha (1.75±0.58), F = 6.839, p = 0.002 (see Figure 2). Mean litter cover was generally within the middle category (33–66%) in the Kararacha and Jilore blocks and below the lower category (0–33%) in the Narasha block (F = 9.937, p = < 0.001).\n\nThe figure shows the comparative depths of forest floor litter across the three forest blocks with the deepest litter in Kararacha and the least in Narasha. The figures represent mean values of litter depth across booth transects in each block. Error bards denote 95% confidence intervals.\n\nOther significant spatial variations in means of habitat variables were observed in overall tree removal (total cut stems), removal of small poles (small-sized trees), and density of live mid-sized trees and removal of mid-sized trees (Table 2). Thus overall tree removal rate was highest in the Kararacha block and lowest in Narasha both for small poles and large mature trees. The same pattern was observed for the density of mid-sized live woody vegetation.\n\nTree removal and live tree figures are given in densities per hectare.\n\nOverall, the Brachystegia habitat was dominated by small-sized trees of 30 cm diameter at breast height (dbh) or less especially in the Jilore area (Table 3). These were also the most intensely logged tree sizes with most of them cut in the Kararacha block (Table 2).\n\nVegetation density is expressed as mean number of live woody stems and logging intensity as mean number of cut stems.\n\nSokoke Pipit abundance was strongly correlated to forest floor litter depth (R2 = 0.719, β = 0.848, p = 0.033) and floor litter cover (R2 = 0.769, β = 0.877, p = 0.021) although litter depth was the better predictor of the species’ abundance (Figure 3), with a predictive equation:\n\nSokoke Pipit abundance = 0.727 + 0.485 * Mean litter depth\n\nThe scatter plot illustrates an overall positive influence of litter depth on abundance and distribution of Sokoke Pipit across the three forest blocks. Abundance is expressed in density per hectare and mean litter depth determined in centimeters.\n\nFurthermore, litter depth was positively correlated to logging intensity of small trees (R = 0.787, p = 0.063) suggesting that pruning of small trees in the forest by tree poachers might be a significant source of forest floor litter. Sokoke Pipit density appeared adversely affected by overall logging intensity (R2 = 0.663, β = -0.814, p = 0.048) see Figure 4. However, there was no significant effect of percent canopy cover (R = 0.5798, p = 0.228) or canopy height (R = 0.174, p = 0.742) on Sokoke Pipit density.\n\nThe figure shows the net impact of tree removal intensity on Sokoke Pipit abundance in the forest. The Sokoke Pipit was encountered less often in areas with high tree loss, which represents degradation of its habitat through understory opening, reduced forest floor litter and possible exposure to predation.\n\n\nDiscussion\n\nThe density of the Sokoke Pipit from this study are lower than the values from studies in the same habitat about a decade ago in which the undisturbed Brachystegia forest had 2.8 birds ha-1 and disturbed zones had 0.9 birds ha-114. The same applies for the previous estimated total population of 13,000 birds. This is because of the continued degradation and modification of the species’ habitat in the Brachystegia spiciformis zone through disturbance, especially in the form of tree cover loss, which has continued over the past decade as observed by many investigators9,11,15,17–20,38. Human activity and related encroachment effects are strongly presumed by all these investigators as the sole and direct source of the disturbance. The results of the present study confirm this but suggest that other complementary causes could be responsible for this habitat degradation processes.\n\nTo put this into perspective, it is important to understand the current general categorization of two main regions within the study area (the woodland dominated by Brachystegia spiciformis) as used by ecologists and forest managers. The first region is the middle section known as Narasha, which is generally characterized as “disturbed” while the second region is Kararacha to the south east, which is characterized as “undisturbed”11,18–20. The Jilore region, which is located to the far northwesterly end of the ASF across the larger Cyanometra forest stand, is rarely studied and is generally uncharacterized using such criteria even though it harbours Brachystegia tree stands. These regions are labelled solely on the basis of the comparative intensities of decades of intensive selective logging, which was spurred by high demand for valuable timber species, leading to loss of much of the primary indigenous stands of Brachystegia trees during the forest’s pre-protection era11,18. The effect of this logging impacted the Brachystegia forest so heavily that it is yet to recover its original near-pristine forest status. It is this readily apparent scar of differential or selective “disturbance” between these regions that still guides most scientific habitat stratification in comparative study design. But this characterization is driven by such visible vegetation structural evidence as comparative densities, extent of spatial cover or openness of the understorey11,17,20,26 rather than the more functional and ecologically consequential processes and dimensions such as species or community dynamics, inter-trophic interactions, effect of anthropogenic encroachment, forest management systems or even climate change.\n\nFor instance, not only is earlier intensive deforestation still discernible in the structure of much of the forest, but in addition human populations around the forest have grown steadily and rapidly over the years9,18. The increasing dependence of these adjacent communities on forest products have resulted in significant negative ecological impacts on the forest’s biodiversity9,14,17,20. In addition, forest management intervention methods by the Kenya Forest Service (KFS) and Kenya Wildlife Service (KWS), have been in place for more than three decades since the era of official intensive logging began and this has further influenced overall ecological processes in the forest including the increasing population of elephants18,25. Partly due to this forest surveillance, much of the illegal logging in the ASF is now predominantly focused on smaller trees or poles that are easier to cut and remove from the forest. Thus, effective characterization of the forest into spatial zones should therefore account for both structural and functional elements in the ASF.\n\nIn this study, the main driver of Sokoke Pipit habitat degradation was tree removal that results in opening up the understorey, a process that may expose individuals to the risk of predation through increased edge39,40, reduction in patch substrate14,26 or change in micro-climate39. One of the main reasons for lower logging rates in Narasha is that it is closest to the KWS and KFS stations and thus enjoys higher levels of surveillance against tree poaching compared to Kararacha and Jilore where logging rates were higher. These patterns conform to patterns observed by Ngala and Jackson from surveys carried out in 2009 and 201025,41. Secondly, it is the region with the highest elephant activity38 which is a further deterrent to illegal loggers.\n\nHowever, the effect of tree removal on Sokoke Pipit abundance was offset by the positive influence of forest floor litter cover and depth. Floor litter harbors much of the arthropod and other invertebrate biomass on which many insectivorous birds such as Sokoke Pipit depend15,42. Secondly, the process of removing small trees appeared to be a significant additional source of floor litter, due to cumulative layers of discarded leaves and twigs left behind by tree poachers during pole harvesting, in addition to the slow rate of decomposition of organic matter typical of many forests along the eastern coast of Africa43,44.\n\nThe Jilore block’s predominance in Sokoke Pipit encounter rates, in spite of its proximity to human settlements and farmland, may in part be due to its comparatively small size and low canopy with an understorey dominated by small regenerating trees (Table 3). Harvesting of small trees for poles, which was highest in this block might also contribute to litter density that is conducive for harboring the Sokoke Pipit’s arthropod prey. On the other hand, the Kararacha block’s high Sokoke Pipit abundance (Table 1) in spite of high logging rates indicates that human-driven selective tree removal is not the sole determinant of Sokoke Pipit population abundance or distribution across the Brachystegia habitat. For Narasha block, low litter depth coupled with lower percent canopy cover and low overall tree density due to poor regeneration all contributed to relative non-suitability for the Sokoke Pipit, despite apparent intensive surveillance against tree logging due to its proximity to the KWS and KFS stations. This is in contrast to Kararacha block’s comparative habitat suitability with respect to these variables is evidenced not only by higher a Sokoke Pipit density but also greater overall birds species richness despite a comparatively higher degree of logging pressure targeting small poles.\n\nEvidence of the role of elephants in modifying the forest habitat is borne by our numerous direct chance observations across the study area, particularly in the Narasha block during which we made frequent sightings of trees felled or broken and the ground dug up by elephants. Analyses of data (see Data File) from these incidental observations was not attempted since counts were made only for the Narasha and Kararacha blocks. However, the spatial distribution of elephant damage noted here is consistent with similar earlier studies and observations conducted by ASFMT18, Ngala41 and Banks et al.38, all of which recorded the highest elephant activity intensity in Narasha. Such intensive activity results in more open canopy, exposed understorey and an increased area of edge habitat that may limit the dispersal capability of species that avoid crossing gaps, or may increase rates of general or nest predation45,46.\n\nTwo main reasons support the contribution of elephants to habitat modification in the ASF’s Brachystegia woodland. First, the forest is estimated to hold between 126 and 184 individuals, giving a density of 0.44 animals km-129,30. Not only does this make the ASF the 7th highest elephant density site of all 30 elephant habitats across Kenya29 but is also fast approaching the 0.5 km-1 recommended maximum carrying capacity, to ensure stability and sustainability of the vegetation in the habitat47. This density is a conservative estimate as it represents a projection for the whole forest; considering that the elephants seem to favour the Brachystegia forest zone38, the carrying capacity will likely be exceeded much sooner than for the ASF overall, with negative consequences for the Sokoke pipit for which this is a critical habitat.\n\nSecondly, the forest management has began erecting an electric fence, in 2006, which already covers a substantial portion of the forest boundary for the purpose of keeping the animals within the forest to reduce conflicts with forest-adjacent farmers who have previously incurred heavy crop losses to the elephants. This physical barrier to dispersal has had the effect of nearly doubling elephant density in the forest, further stretching the carrying capacity and worsening the habitat degradation process47. The pressure is particularly high in the ASF due to its small size in comparison to other elephant sites in Kenya30 and given the peri-urban nature of the forest with its surrounding agricultural land and human settlements18.\n\nIn addition, the Brachystegia vegetation zone of the ASF has the lowest vegetation regeneration rates along the entire eastern coast of Africa due to soil with a functionally poor structure24, low nutrient content, low moisture level and limited micro-organism activity that is necessary for nutrient cycling11,44.\n\nThus, in terms of forest habitat health, in addition to human-driven tree removal the high elephant density and the restrictive nature of the electric fence are compounded by the slow forest regeneration rate, which may aggravate the role of elephants as drivers and accelerators of overall habitat change in the Brachystegia woodland. In many sections along the transects in the Narasha area, the frequency of elephant-felled trees outnumbered those cut down by humans. Without implementing measures to regulate the elephant population and movement in Brachystegia woodland concurrently with a halt in illegal logging in the forest, the rate of tree removal is likely to increase in the medium term with serious ramifications for the Sokoke Pipit and other forest-dependent species.\n\n\nConclusions\n\nThe Sokoke Pipit’s favoured habitat is an open understory with deep litter cover, often but not always with dense vegetation. Its density and estimated population in Brachystegia woodland is lower than it was a little more than a decade ago, suggesting increased pressure on the species’ habitat through increased loss or continued modification. The main cause of this habitat degradation is still illegal logging. Habitat degradation may be further accelerated by elephant–mediated habitat damage through tree felling, as evidenced by several incidental observations during the study, and consistent with findings of other recent studies though more in-depth studies on this are needed to underpin its scale and impact patterns on Sokoke pipit and forest specialist birds. Tree poachers mainly target small trees/poles which are taken mainly from parts of the forest farthest away from patrol bases and with minimal elephant numbers, from where they are easily carried out of the forest. A sound long-term conservation strategy would involve significantly reversing tree logging trends through increased surveillance; effectively managing local elephant populations and their movement; and stepping up habitat restoration through reforestation of heavily damaged Sokoke Pipit sites.\n\nThe deterring effect of proximity of Narasha to the KWS and KFS stations indicates the potential benefits of increased patrol and surveillance as an immediate check on habitat destruction by humans not only in the Brachystegia woodland zone but also throughout the forest.\n\n\nData availability\n\nfigshare: Arabuko-Sokoke forest ecological data: Sokoke Pipit abundance, vegetation survey results, floor litter measures and elephant damage in three forest blocks, http://dx.doi.org/10.6084/m9.figshare.92469048.", "appendix": "Author contributions\n\n\n\nNO conceived the study and designed the experiments, NO prepared the first draft of the manuscript while NO, DN and AM were all involved in the process of project planning, logistical arrangements, data collation, data summary and revision of the initial project report. They all 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\nFunds for the project were kindly provided by The African Bird Club through its Conservation Programme. Funds were awarded to NO in 2011. Additional financial and logistical support was kindly provided by the National Museums of Kenya.\n\n\nAcknowledgements\n\nWe greatly thank the Kenya Wildlife Service and Kenya Forest Service for permitting us to carry out the study in the ASF; the Kenya Forestry Research Institute Coastal Eco region for allowing us access to reference material; the Arabuko-Sokoke Forest Guides Association for recommending and allowing participation of the two members. A project report version of this article was originally posted on the African Bird Club website: http://www.africanbirdclub.org/sites/default/files/2011_Sokoke_Pipit.pdf.\n\n\n\n\nThe checklist is in phylogenetic order grouping birds by family, scientific and common name following Bird Committee of the east African Natural History Society49.\n\n\nReferences\n\nGardner TA, Barlow J, Chazdon R, et al.: Prospects for tropical forest biodiversity in a human-modified world. Ecol Lett. 2009; 12(6): 561–582. PubMed Abstract \| Publisher Full Text\n\nWhitmore TC: Tropical forest disturbance, disappearance, and species loss. in Laurence WF and Bierregaard RO. (Eds) Tropical Forest Remnants, Chicago University Press, Chicago. 1997. Reference Source\n\nCollar NJ, Crosby MJ, Stattersfield AJ: Birds to watch 2: The world list of Threatened birds. BirdLife International, Cambridge, UK. 1994. Reference Source\n\nStattersfield AJ, Crosby MJ, Long AJ, et al.: Endemic bird areas of the world: priorities for bird conservation. BirdLife International, Cambridge, UK. 1998. Reference Source\n\nBennun LA, Howell K: Bird sampling techniques. In: African Forest Biodiversity: A Field Survey Manual for Vertebrates. G. Davies ( Ed). Earthwatch Institute (Europe), Cambridge, UK. 2002; 131–161. Reference Source\n\nVan Someren VGL: Eremomela badiceps turneri. subsp. nov. Bulletin of the British Ornithologists' Club. 1920; 40: 92.\n\nBurgess ND, Cutts CJ, Huxham M: New records of the Sokoke Pipit Anthus sokokensis from Kiono Forest Reserve, Bagamoyo District, northeastern Tanzania. Scopus. 1991; 15: 56–57.\n\nDowsett RJ, Forbes-Watson AD: Checklist of birds of the Afrotropical and Malagasy regions. Liège, Belgium: Tauraco Press. 1993. Reference Source\n\n9 BirdLife International. Species factsheet: Anthus sokokensis. 2013. Reference Source\n\nInternational Union for Conservation of Nature: IUCN Red List of threatened species. Version 2012. 2013. Reference Source\n\nOyugi JO, Brown J, Whelan CJ: Effects of human disturbance on composition and structure of Brachystegia woodland in Arabuko-Sokoke Forest, Kenya. Afr J Ecol. 2007; 46(3): 374–383. Publisher Full Text\n\nFry CH, Pearson DJ, Taylor PB: Motacillidae: Wagtails, Pipits and Longclaws. In: Keith S, Urban EK and Fry CH (Eds). The birds of Africa Vol. IV. Academic Press Ltd. London (1992).\n\nMlingwa COF: A note on the rediscovery of the Sokoke Pipit Anthus sokokensis in the Pugu Hills, Tanzania. Bird Cons Int. 1996; 6: 293–294. Publisher Full Text\n\nMusila F, Bennun LA, Karanja W: The Sokoke Pipit Anthus sokokensis, in Arabuko-Sokoke Forest, Kenya: population estimates and response to habitat disturbance. Ostr Suppl. 2001; 198: 97–204.\n\nBennun LA, Njoroge P: Important Bird Areas in Kenya. East African Natural History Society, Nairobi. 1999. Reference Source\n\nWaiyaki EM, Bennun LA: The avifauna of coastal forests in southern Kenya: status and conservation. Ostrich. 1999; 71: 247–256. Publisher Full Text\n\nMatiku PM, Bennun LA, Nemeth E: Distribution and population size of the threatened East Coast Akalat in Arabuko-Sokoke Forest, Kenya. Ostrich. 2000; 71(1-2): 282–285. Publisher Full Text\n\nArabuko-Sokoke Forest Management Team: Arabuko-Sokoke Strategic Forest Management Plan 2002–2027. Birdlife International, UK. 2002.\n\nOyugi JO, Brown JS, Whelan CJ: Foraging behavior and coexistence of two sunbird species in a Kenyan woodland. Biotrop. 2011; 2: 1–8. Publisher Full Text\n\nDavis JA: Density and population estimates of Amani Sunbird Anthreptes pallidigaster in Kenya’s Arabuko-Sokoke Forest. Bird Cons Int. 2005; 15: 53–62. Reference Source\n\nBurgess ND, Doggart N, Doddy K, et al.: New information on the lowland coastal forests of eastern Africa. Ory. 2003; 37: 280–281. Reference Source\n\nGordon I, Burgess N, Luke Q, et al.: Final project report of the Critical Ecosystem Partnership Fund on the Easter Arc Mountains and coastal forests of Kenya and Tanzania: ecosystem profile. Conservation International, Cambridge. 2003. Reference Source\n\nCollar NJ, Stuart SN: Key forests for threatened birds in Africa. ICPB Monograph No. 3. ICPB, Cambridge. 1988. Reference Source\n\nUNESCO: The Eastern Arc Coastal Forests: Arabuko-Sokoke forest and Shimba Hills National Reserve. UNESCO World Heritage Centre. Fontenoy, France. 2013. Reference Source\n\nNgala D, Jackson C: Monitoring Important Bird Areas in the Arabuko-Sokoke Forest. Friends of Arabuko-Sokoke Forest. Malindi, Kenya. 2010. Reference Source\n\nFanshawe JH: The effects of selective logging on the bird community of Arabuko-Sokoke forest. Unpublished PhD thesis, University of Oxford, UK. 1995. Reference Source\n\nKenya Commission on Revenue Allocation: County Fact Sheet: Kilifi County. Government of Kenya. Nairobi. 2012.\n\nMuoria PK: Ecological correlates of crop raiding by elephants and baboons: a case study in Arabuko-Sokoke Forest, Kenya. Unpublished PhD thesis. Kenyatta University, Kenya. 2001. Reference Source\n\nOmondi P, Litoroh M, Kock R, et al.: Conservation and Management Strategy for the Elephant in Kenya (2012–2021). Kenya Wildlife Service. Nairobi. 2012. Reference Source\n\nInternational Union for Conservation of Nature: Elephant Database: Kenya 2012 (Africa Analysis). 2012. Reference Source\n\nBibby C, Jones M, Marsden S: Bird Surveys: Expedition Field Techniques. London: Expedition Advisory Centre (Royal Geographical Society). 1998; 143. Reference Source\n\nSutherland WJ Ed: Ecological census techniques: A handbook. Cambridge University Press, Cambridge. 1996. Reference Source\n\nBuckland ST, Anderson DR, Burnam KP, et al.: Distance Sampling. Estimating the Abundance of Biological Populations. Chapman and Hall, London. 1993. Reference Source\n\nFewster RM, Buckland ST, Burnham KP, et al.: Estimating the encounter rate variance in Distance sampling. Biometrics. 2009; 65(1): 225–236. PubMed Abstract \| Publisher Full Text\n\nZar JH: Biostatistical Analysis. 4th ed. Prentice Hall, New Jersey. 1999. Reference Source\n\nMarques TA, Thomas L, Steven GF, et al.: IMPROVING Estimates of bird density using multiple covariate distance sampling. The Auk. 2007; 124(4): 1229–1243. Publisher Full Text\n\nBuckland ST, Anderson DR, Burnham KP, et al.: Advanced distance sampling: estimating abundance of biological populations. Oxford University Press, Oxford, UK. 2004; 400. Reference Source\n\nBanks J, Jackson C, Hannon L, et al.: The cascading effects of elephant presence/absence on arthropods and an Afro-tropical thrush in Arabuko-Sokoke forest, Kenya. Afr J Ecol. 2010; 48(4): 1030–1038. Publisher Full Text\n\nLindenmayer DB, Fischer J: Habitat fragmentation and landscape change: an ecological and conservation synthesis. Island Press. Washington. 2006. Publisher Full Text\n\nNewmark WD, Stanley TR: Habitat fragmentation reduces nest survival in an Afrotropical bird community in a biodiversity hotspot. Proc Nat Acad Sci U S A. 2011; 108(28): 11488–11493. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNgala D: Monitoring tree poaching & elephants in Arabuko-Sokoke Forest with Kenya Forest Service. Unpublished technical report of Friends of Arabuko-Sokoke Forest. Malindi, Kenya. 2009. Reference Source\n\nLange CN: Environmental factors influencing land snail diversity patterns in Arabuko Sokoke forest, Kenya. Afr J Ecol. 2004; 41(4): 352–355. Publisher Full Text\n\nVan-Aarde RJ, Ferreira FM, Kritzinger JJ, et al.: An evaluation of habitat rehabilitation on coastal dune forests in northern KwaZulu-Natal, South Africa. Rest Ecol. 1996; 4(4): 334–345. Publisher Full Text\n\nBackéusa I, Pettersson B, Strömquist L, et al.: Tree communities and structural dynamics in Miombo (Brachystegia–Julbernardia) woodland, Tanzania. For Ecol Manag. 2006; 230(1–3): 171–178. Publisher Full Text\n\nBender DJ, Contreras TA, Fahrig L: Habitat loss and population decline: a meta-analysis of the patch size effect. Ecology. 1998; 79(2): 517–533. Publisher Full Text\n\nNewmark WD: Tropical Forest Fragmentation and the Local Extinction of Understory Birds in the Eastern Usambara Mountains, Tanzania. Cons Biol. 2005; 5(1): 67–78. Publisher Full Text\n\nJachmann H, Croes T: Elephant effects on woodlands and a suggested optimum elephant density at the Nazinga Game Ranch, Burkina Faso. Unpublished technical Report No LAEBRA 1989/142 of Association de Dévelopement de l’Élevage de la Faune Africaine. Ougadougu, Burkina Faso. 1989. Reference Source\n\nNickson EO, Ngala D, Mwalimu A: Arabuko-Sokoke forest ecological data: Sokoke Pipit abundance, vegetation survey results, floor litter measures and elephant damage in three forest blocks. figshare. 2014. Data Source\n\nBird Committee of East African Natural History of Society: Checklist of the Birds of Kenya, 4th Ed. EANHS, Nairobi. 2009." }	[ { "id": "3968", "date": "14 Mar 2014", "name": "Jeremy Lindsell", "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\nOverall comments:This is an important survey, and I support the collection of these results, but not the way they are currently presented and interpreted. The presentation has been compromised by poor attention to methods. This raises concerns about errors in the analysis and that conclusions are not assured. The small sample size involved also raises concerns about the robustness of the conclusions. Given the very large population decline being described it is important to ensure that the conclusions are supported by the data. As it stands a lot more caution needs to be attached to this work - results might be suggestive of a decline that needs more serious investigation rather than this being proven. Furthermore, explanation of the declines is not always clear – the overall tone of the discussion is that disturbance has harmed the pipit even though some evidence suggests the pipits benefit from disturbance. No data are presented to demonstrate that disturbance has increased over the course of the decline. I think the authors could do with a good session with a qualified statistician to first straighten out the analyses, and then review their conclusions. And I would urge them to do this as the conclusions could be important for the conservation of this species. Abstract:5 globally threatened and 4 near-threatened, not 9 threatened. The Methods section states it correctly. Introduction:Birds in general are not most sensitive to forest change, but certain groups are good indicator species. Avoid reference to BirdLife factsheets if possible, good as they are - the original source is preferable. The title of the article by Musila et al. (2001) certainly implies they studied response to habitat change. Make clear that you are not studying change – it is not a before and after study. You are substituting space for time. The title also implies you have studied change over time (modification). Materials and methods:Change “BirdLife protocols” to “BirdLife criteria”. The second paragraph is largely not needed as it is unrelated to the pipits, and can be referenced elsewhere. Third paragraph, change “community zones” to “communities”. Paragraph four – insert a quick statement about the amount of forest in the surrounding landscape (i.e. almost none!) How do honey harvesting and game snaring affect the habitat? Do they cut trees to harvest honey? But what about snaring? Sampling strategy:You use the term “lumbering”. If this is to denote commercial logging as opposed to illegal logging might be better to make this clear. The positioning of transects is very unclear. Are you saying that you used existing paths as transects? Or were the paths only used to locate a starting point? You also need to say why you used existing paths as transects when you have already said that the habitat has a relatively open understory and is therefore presumably easy to move through. Looking at the map, the transect look very straight which suggests they were not on existing tracks. Do you think the 3rd track on the left really is unbiased since more heavily disturbed areas are likely to have a higher density of tracks, which means your sample in more disturbed areas will always tend to be nearer your point of entry and likely therefore to be the more disturbed parts of the patch. I think you need to be clear about this. You say you ensured at least 1 km separation between transects but don’t say how you actually did this given that the transect selection is described as random. Bird Survey:Distance sampling (as cited) does not specify a fixed width to the transect. It is called “variable distance” sampling. You may specify a cut off beyond which you don’t bother recording observations. Is that what you mean? Doubling counting of birds on different transects is only a problem if the movement of the bird between the two was caused by the observers. Presumably you recorded cluster size? It isn't mentioned. Did you walk transects more than once? This is implied in the paragraph on sampling strategy but no information is given about how many times here or there. Vegetation sampling:How did you measure % canopy height with rangefinder? How many readings did you take and how do you express this as a percentage? Percentage of plot covered by different height classes? If you measured straight up this tends to underestimate height because of the number of occasions that the laser hits lower branches and because you only read from the underside of the canopy. If you measured from the distance then you needed to measured angle too? Was the stem size you measured the circumference or the DBH. You say you used an ordinary tape measure. Why do you say this - you can still use this to record diameter. Are the size classes given circumferences or diameters? You say you recorded cut stems in “similar” size classes to the live stems. What do you mean? The same? If not the same say why and how the classes differed. Obviously you can’t usually measure a cut stem at breast height so need to say so and how you dealt with this. In that sentence about cut stems you also refer to diameter – so does that mean the other measurement earlier were diameters? Floor Litter Sampling:“Litter cover was assessed by dividing the 10 × 10 m quadrats into 25 smaller grids of 2 × 2 m quadrats by use of a standard meter rule and tape measure then ascertaining the percentage category in each 2 × 2 square before averaging the total out of 25” This is an awkward sentence. Did you ascertain the percentage cover in each 2x2 square and then record the average value for all 25 squares? Data analyses:Try to avoid starting section/paragraph/sentence with “because”. OK, so here you state the number of visits – this is in the wrong section. You shouldn’t being expecting normal distribution in bird count data or counts of trees in stem size classes, regardless of samples sizes. Counts are discrete integers whereas a normal distribution assumes all possible values could be observed (i.e. fractional values). This means that log-transforming the count data to cope with the apparent skew in the data is incorrect. Count data in natural populations are nearly always heavily skewed – for birds there are always lots of low counts (and often many zero counts) and for trees there are (nearly) always lots of small trees compared to big trees. Bird count data often follow a poisson distribution. Having said all that there shouldn’t be any transformations undertaken of the bird data (the distances) prior to analysis in Distance. You are not analysing the counts but the distances and in this case you certainly wont have a normal distribution – at best a half normal. The distribution of predictor variables like stem density is not important and don’t need transforming for a regression. But if you are analysing stem density by forest block then stem density becomes the response variable and you do need to worry about its distribution. When you analyse using regression the densities arising from the Distance analysis then you need to worry about the distribution. And in this case you may need to transform the densities to approximate normality. For calculation of the encounter rate, “n” is not “mean abundance”, it is the “number of detections”. You are using the number of detection to try and work out the abundance. Explanation of choosing MCDS isn’t clear. If cluster size was entered as a covariate because cluster size affected detectability then that makes sense, but you state that cluster size varied little implying it wasn’t an issue. But you do seem to say that encounter rate varied a lot from one site to another. But you’ve not made it clear why MCDS would help cope with this (actually I don’t think it would). “We selected the cosine adjusted half-normal detection functional model with the lowest value based on Akaike Information Criterion in the density estimations“ would read better as: “We selected the cosine adjusted half-normal detection function based on it having the lowest AIC value”. However, you ought to assess model fit on more than just AIC. Did you have to pool data? Or truncate data? What other models did you consider? “Bird diversity was worked out using the reciprocal of Simpson’s…” better worded as “Bird diversity was estimated using the reciprocal of Simpson’s…” It is not clear the level at which stem densities were calculated – for each transect? Same goes for the other veg measurements. IT is not clear what you did to canopy cover: converted the values to ordinal scale then converted this to a ratio. Ratio of what to what? Doesn’t this imply a closer numeric relationship between the classes than you really measured? Presumably the high variance in stem densities was because of the very small plot size relative to mean stem density – 0.1 ha is quite a small area to sample for trees in any forest. “simple linear regression” - It is not clear what this means. Do you mean univariate tests where you only consider one predictor variable at a time? If so, then you really ought to consider multivariate tests. “Differences of means of the key habitat (independent) variables were compared on the spatial scale by one-way Analysis of Variance (ANOVA) using the forest blocks as the categorical treatment effects on the bird (response) variables.” I don’t understand this sentence so please rephrase. Are you saying you used ANOVA to test for differences in the habitat variables between the different forest blocks? Not sure how the “bird (response) variables” also fit in here. Results:No need to report on other birds species. The paper is about the pipit. You can mention that pipits surveyed as part of survey of all species but the other species data are few and add little. “There were 17 encounters of Sokoke Pipit with an overall abundance of 30 individuals” Confusing. You encountered pipits on 17 occasions comprising 30 individuals. Abundance is what you are trying to estimate from these encounters. You could put mean group size in brackets here to clarify further). Jilore and Kararacha are described as “moderate to highly disturbed” in contrast to Narasha which is “more disturbed”. Not clear which is actually the more disturbed. Table suggests the former are less disturbed. You report results of a chi sq test of clumpedness but don’t explain this in methods. Can you report 95% confidence intervals instead of standard errors in table 1. Makes it much easier to interpret the results. Table 1 says you right-truncated the data – presumably you mean in Distance. This is not mentioned in Methods. Perhaps relates to earlier confusion about appearing to use fixed width transects at 60m. Table 1 AIC presumably refers to the value returned by distance. This serves no purpose in this table as you cannot compare AIC values for the different forest blocks as they are computed from different data. AIC values can only be compared where the data are the same. Is degradation confounded with edge effect (Jilore)? Figure 2 - there should be no line connecting the three sites because there are no intermediate locations to be represented by positions along the line. They are discrete sites. Would be better to have a box and whisker plot for this figure. Fig 2 - it is hard to believe from this plot that litter depth does vary significantly between the blocks despite the reported very low p value! Table 2 should say stems per hectare, not densities per hectare. Having said stems size classes were pooled earlier on, table 2 reports results of them unpooled. Why can’t Fig 4 be shown as scatter plot like Fig 3? There appears to be no relationship in Fig 4 between logging and pipits. Two unlogged sites have both high and low pipet density and the logged site had intermediate pipit density. Good to include the raw Distance data – that’s commendable. Would be also good to have a plot showing the Distance histogram and selected model to demonstrate how suitable the selected model was. The analysis really lacks a coherent conclusion as there was no attempt to combine the predictor variables in a single analysis. Admittedly this may not be possible with so few replicates but there may be other ways to cut the data to improve this. Discussion: A good edit for language is required, in the discussion especially. There are long sections about forest disturbance and history that do not arise from the results of the survey. These should be moved to the introduction by way of describing the site and why the survey might be needed or removed. Not sure that the impact of logging on the pipits is proven from the presentation given. Fig 4 is not convincing as it stands. In fact the opposite seems to be argued for in places – logging leads to more litter which is good for pipits. The article needs a good reference to back up the idea that leaf litter from felled trees persists for any length of time. Usually if a tree is felled the leaves dry up attached to the branches and don’t fall off as they would naturally. Logging may lead to there being more dead wood rotting on the forest floor though. The elephant issue is important as fenced population must be having an impact. My understanding was that the fence was now complete. There is no real discussion of comparison with former surveys. Decline is rather simply attributed to habitat degradation without comparing the habitat in current survey with the habitat before.", "responses": [ { "c_id": "747", "date": "01 Apr 2014", "name": "Nickson Otieno", "role": "Author Response", "response": "Dear Dr. Lindsell,We highly appreciate your review comments on our article, particularly the accompanying detailed suggestions for revision. We have now gone through the comments and revised the article accordingly by addressing the concerns as outlined below, and hope that our treatment meets with your expectations.Thanks,Nickson E. Otieno (for the co-authors) Revisions by the authors in detail:AbstractWe have now corrected the proportion of globally threatened and near-threatened species.IntroductionWe have clearly specified that the group of birds useful for monitoring forest health are the forest-dependent ones. We have now minimized repeat references to BirdLife Fact sheets. We have clarified that our study, in contrast to that of Musila et al. which examined species’ response to habitat change, dealt also with spatial variations in habitat structural quality, and that the results provide an update in the spatial and temporal dimensions of the habitat effects on the species. Materials and methods:We have changed BirdLife protocols to BirdLife criteria as suggested. We have retained paragraph two, as we feel it presents a suitable background the forest as a significant national and global habitat for many forest-dependent species of which SP is one. But we have eliminated the part outlining other taxa found in the forest. We have changes community zones to communities as suggested. We have included a statement about the absence of any forest fragment within the agricultural zone outside the main forest blocks. We have clarified that honey harvesters and game hunters damage the forest habitat by clearing vegetation to make paths, burning vegetation to access hives. Creation of such openings further accelerated anthropogenic impacts on the forest habitat, which affects forest-interior species like SP. Sampling strategyLumbering: we have clarified that this was officially sanctioned commercial logging, in the era before forest protection was actively enforced.Transect positioningWe did not use existing paths, as you rightly point out, the habitat was penetrable enough for survey walking, and the paths were only use for purposes of location the starting points of transects. On the map the transects look straight but in actual fact they were not necessarily so. The illustrations representing the transects on the map show only the straight lines joining the starting and ending points of transects. We have included this clarification in the main text. Using the 3rd track to select the start of transects was considered sound enough as a way of randomizing transects across the blocks because despite there being differences in disturbance levels across the blocks, presence of the paths did not differ significantly across them (see results, final paragraph) in frequency and thus was unlikely to be a significant source of bias. The “3rd track” tool was thus just a way of systematizing transect randomization to minimize spatial bias. It is this systematization scheme that ensured that if we could not maintain at least 1 km between adjacent transects, we continued searching till we found a transect starting point that fulfilled both conditions. So the abundance of tracks in the forest indeed assisted in some way. Bird surveyWe have specified that we used Distance sampling but “fixed” our maximum transect width to 60m, beyond which, even in a forest that is not exactly too thick, it makes it subjective to accurately detect all individuals or clusters of a species as sensitive, silent and camouflaged as the SP. For this reason we also made truncation of our distance values on the progmamme itself by a general value of 5m. It is also clearly mentioned that we recorded SP individuals “and clusters” because there were incidents in which only single individuals were detected/encountered. We have added that bird transects were run twice each on different days. Vegetation samplingIt is canopy cover that was determined into % and not canopy heights. The canopy heights were determined using a range finder not from the within each transect quadrat but from an open area, either on a track or a deforested area, and using the range finder to obtain the observer distance from the tallest crown tree and then measuring to the crown height then using trangulation to determine crown height and adding eye-level height. We have clarified that live stems and cut tree stumps were measured in terms of circumference (not diameter) size classes. Only live stems were measured at breast height.Floor litter samplingThe sampling description is now made clearer. Data analysisThe count data that was log transformed initially were those of other counted things as only live stems and logged-out tree stumps. Bird data was only transformed from encounter rates (abundance) for purposes of regression against litter depth. By “transformed into densities per hectare” with regard to stem densities, we actually means “expressed as densities per hectare”. We have now reflected this in the data analysis section text. The initial transformation by logarithm was for the purpose of comparing these variables between blocks. We have clarified that SP encounter rates were worked out from total number of detections divided by survey effort, which was 2 km of transect in each forest block, surveyed twice each. Accordingly, we have reworked the encounter rates and corrected the values. It was our feeling that the rather small sample size of pipit detections was related to disturbance effects on the Pipit habitat within the brachysetgia forest. Having shown that disturbance as a parameter itself varies across the three segments of Brachystegia forest, we felt that it would significantly influence detection of the pipits. One evidence of the variant disturbance across the blocks was the mean sighting/detection distance, which also ended up corresponding to the forest disturbance levels of the blocks. Therefore MCDS was employed by using disturbance levels as factor covariates thus reducing the variance (and possible low confidence) in the density estimates that would be expected if CDS were to be used. In determining the densities using Distance, data were pooled for the various blocks into a global analysis, factoring in the block factor covariates. But the densities were also worked for the individual blocks. For the global analyses, we have also replaced the standard errors of density estimate on table 1 with 95% confidence intervals. We have also included a description of how the model of fit was selected for distance estimation, and how data was truncated for analysis. Vegetation assessment variables analysis were treated at the transect and block levels. Percent canopy cover scores were coded such that open canopy, moderately open canopy and closed canopy scored 1, 2 and 3, respectively. These were then transformed to ratios scaled with ‘3’ as the maximum. So the ratio was cover score:3. Transformation of the ratios using ArcSine ensured that there would be no close relationship between the rations representing the cover scores than was actually measured. By simple linear regression, we mean “neither logistic nor loglinear”. Multivariates was the method, through which for instance were selected litter depth as a better predictor of pipit abundance than litter cover as we had stated in the results section. Although 10 x 10 m quadrats could be small for sampling forest trees, our analyses of vegetation measurements were done at the transect level which integrated 10 of the qudrats of each transect thereby reporting results per ha rather than 0.1 ha. The statement on ANOVA means that means were compared across the blocks. We have revised the statement to read: “Means of habitat variables were compared across the blocks using one-way ANOVA” ResultsThe section mentioning other bird species has been removed in the revised version of results. Jilore and Narasha blocks are described as less disturbed as compared to the more disturbed Narasha block. We have included in the data analysis section the use of the chi test for S Pipit distribution. It is possible that edge effects could be linked to effects of degradation in Jilore block. However, as we did not investigate extent or effects of edge effect, our main view about the high detection of S Pipit in that block is related to lack of massive destruction by elephants and comparatively reduced human traffic over the past few years due to the enclosing electric fence barrier. AIC values have been removed from Table 1. Figure 2 is now reproduced in box and whisker form to more distinctly show variation in litter depth across the blocks. Table 2 actually presents the tree stem data pooled into size classes (small sized = >30cm, mid-sized = 31-60cm and large = >60cm) Figure 4 is not presented as a scatterplot because the logging data used to produce it are those of total trees stems cut, as the figure shows (human removal) without inclusion of trees removed or felled by elephants. In the discussion, we clearly showed that impact on Sokoke pipit due to habitat degradation was both a function of stem cutting as well as elephant tree removal. Furthermore, the figure as presented is intended to demonstrate that a slight increase in tree cutting/removal can correspond to a drastic impact on the Pipit abundance. Accordingly Narasha with low logging rate also had low pipit encounter rates, because the habitat degradation in that block is due to the numerous elephants rather than from human-mediated logging (stem cutting). We have updated the legend for Fig 4 to reflect this clarification. DiscussionIt is our view that the section that deals with description of many authors’ characterization of Arabuko Sokoke forest as a way of delineating it in terms of disturbance zones, provides a good setting in which we present our own characterization based on actual observed attributes which are in addition to, rather than restricted to, spatial variations in tree logging patterns. For instance, no other researcher has ever appeared to notice the possible relationship between the elephant feeding habits and forest habitat impacts. Putting this section in the introduction would imply that it is common documented knowledge, which it is not. We have however removed the first paragraph of that section, which might have been the more redundant of earlier descriptions under “Materials and methods”. We were not able to find any study linking Sokoke Pipit needs with habitat variables ever since Musila et al did so in 2000, which is why we did not have much such discussion in our paper. We mentioned however that since the Musila et al study, there has been a decline in Pipit density, presumably due to habitat degradation that has continued since then. This is clearly outlined in the first paragraph of discussion. Oyugi, Fanshawe, Banks, Davis et al. all studied habitat of the Brachystegia forest, but in reference to other species mostly of the forest canopy and thus not directly comparable to the Sokoke Pipit. The impact of tree loss is proven as the main cause of Sokoke Pipit habitat in terms of abundance and distribution and as we argue in the discussion, trees are lost not only through logging by human (fig 4) but also by elephant tree damage. Augmentation of leaf litter (important for pipit) from pruning poached poles in the forest was neither evident throughout the study area nor considered the main driver of Pipit abundance and distribution. It was only associated to areas where logging intensity targeted small trees (human-induced removal). Again, human-induced tree removal was not the only driver of S pipit demographics." } ] }, { "id": "3739", "date": "24 Mar 2014", "name": "John Banks", "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 addresses the links between habitat condition and an endangered bird species in an important forest reserve (ASF) in eastern Kenya. It addresses an important topic, especially given ongoing anthropogenic pressures on this and similar types of forest reserves in eastern Kenya and throughout the tropics. Despite the rather small temporal and spatial extent of the study, it should make an important contribution to bird and forest conservation. There are a number of issues with the methods and analysis that need to be clarified/addressed however; furthermore, some of the conclusions overreach the data collected, while other important results are given less emphasis that they warrant. Below are more specific comments by section:Abstract:The conclusion that human-driven tree removal is an important contributor to the degradation of ASF is reasonable given the data reported in the article. Elephant damage, while clearly likely a very big contributor to habitat modification in ASF, was not the focus of the study (the authors state clearly in the Discussion that elephant damage was not systematically quantified, and thus no data were analyzed) – and thus should only be mentioned in passing here – if at all.Introduction:More information about the life history ecology of A. sokokensis would provide welcome context here. A bit more detail about breeding sites as well as dispersal behavior etc. would be helpful – and especially why these and other aspects render the Pipit a good indicator species/proxy for habitat condition. This could be revisited in the Discussion as links are made between habitat conditions and occurrence of the bird (where you discuss the underlying mechanisms for why it thrives in some parts of ASF and not others, and why it’s abundance correlate strongly with some types of disturbance and not others). Again, you reference other studies that have explored other species in ASF and forest disturbance, but do not really explicitly state why the Pipit is a particularly important indicator of forest condition.Methods:Bird Survey: As described, all sightings and calls were recorded and incorporated into distance analysis – but it is not clear here whether or not distances to both auditory and visual encounters were measured the same way (i.e., with the rangefinder). Please clarify. Floor litter sampling: Not clear here whether or not litter cover was recorded as a continuous or categorical variable (percentage). If not, please describe percentage “categories” used.Results:Mean litter depth graph (Figure 2) and accompanying text reports the means and sd but no post-hoc comparison test (e.g. Tukey HSD) – need to report the stats on which differences were/were not significant. Figure 3 – you indicate litter depth was better predictor of bird abundance than litter cover, but r-squared is higher for litter cover. Need to clarify (and also indicate why you chose only to shown depth values in Figure 3. The linear equation can be put in Figure 3 caption (not necessary to include in text). Figure 4 – stats aren’t presented here; also, the caption states that tree loss and leaf litter are inversely correlated – this might be taken to mean, given discussion (below) about pruning, that there could be a poaching threshold below which poaching may pay dividends to Pipits (and above which Pipits are negatively affected). This warrants further exploration/elaboration. The pruning result is arguably the most important one here – this suggests an intriguing trade-off between poaching and bird conservation (in particular, the suggestion that pruning by poachers may bolster Pipit populations – or at the very least mitigate against other aspects of habitat degradation). Worth highlighting this more in Discussion.Discussion :Last sentence on p. 7 suggests causality (“That is because…”) – but your data only support correlation (one can imagine that there may have been other extrinsic or intrinsic drivers of population decline). P. 8: discussion of classification of habitat types in ASF is certainly interesting, but could be made much more succinct in keeping with focus of this paper. P. 9, top: first paragraph could be expanded – as noted before, tradeoff between poaching/pruning and Pipit abundance is worth exploring in more depth. Could your results be taken as a prescription for understory pruning as a conservation tool for the Sokoke Pipit or other threatened species? More detail here would be welcome (and also in Conclusion); in subsequent paragraph about Pipit foraging behavior and specific relationship to understory vegetation at varying heights could be incorporated into this discussion. Is there any info about optimal perch height for foraging or for flying through the understory? Linking to results of other studies in ASF, is there potential for positive correlations with optimal habitat conditions for the other important bird species in ASF in order to make more general conclusions about management?", "responses": [ { "c_id": "769", "date": "12 Apr 2014", "name": "Nickson Otieno", "role": "Author Response", "response": "Dear Dr. John Banks,Thank you for your useful comments on our manuscript. We have attempted to address your concerns about the manuscript in the following ways:Introduction:We have now added more information about the habits of the Sokoke Pipit, although we have also indicated that no comprehensive previous studies are known of the species’ live history traits, especially breeding records. We have also added that the specialist attributes of the species, especially as the key forest-floor specialist of the interior Brachystegia forest, make it a good candidate for monitoring habitat quality of the Brachystegia understory.Methods:Only once was an encounter of the Pipit based on a call only, and in this case the abundance was assumed to be that of 1 bird and perpendicular distance determined for its approximate location as for the other cases. Litter cover was recorded as percentages of cover in three main categories: fully covered (67-100%); moderately covered (34-66%) and not little or no cover (0-33%). Scoring these respectively as 3, 2 and 1 on a scale of 1 to 3, each was then divided by “3” to derive a cover score that were finally arcsined transformed towards normality of distribution.Results:We have now reported the post hoc Tukey test results statistics for the significant difference in forest floor litter depth across the three blocks. For figure 3 (now figure 4) we have now presented the figure with fresh partial regressions of both the mean litter depth and the arcsines of mean litter cover percent against Sokoke pipit densities per transect. Accordingly, we have revised the legend of this figure to reflect these changes. We have also made the correction in the text of the results section, in which the regression stats for how Sokoke Pipit density varied with litter depth and litter cover were initially written in reversed order. The regression equation between Sokoke Pipit density and litter depth is now transferred from results text body to the caption of figure 4 (formerly figure 3). Actual figures and statistics on logging rate were presented already in Table 2.Discussion:We have now revised the statement that the lower density of the S. Pipit recorded in the study, in comparison to the earlier survey, was “because of” continued habitat to read “attributable to\" instead. The part of the discussion which details the ASF zonation and characterization is now cut down to the essential facts directly relevant to the study. We have expounded from discussion through to conclusion on the issue of the apparent tradeoff between leaf litter from trees pruned by poachers and the quality of Sokoke Pipit habitat, stressing that while such compensatory effects may be beneficial, it might be unwise to recommend understory pruning of young trees, as this might further degrade habitat for S. Pipit and other understory species. Instead, work should be stepped up towards preserving dead wood and controlling forest fires that would reduce forest litter. We have also put recommended conservation measures for the species’ habitat in perspective of recommendations from previous studies in ASF." } ] } ]	1	https://f1000research.com/articles/3-59
https://f1000research.com/articles/3-60/v1	19 Feb 14	{ "type": "Research Article", "title": "An image-based, dual fluorescence reporter assay to evaluate the efficacy of shRNA for gene silencing at the single-cell level", "authors": [ "Shin-ichiro Kojima", "Gary G. Borisy", "Gary G. Borisy" ], "abstract": "RNA interference (RNAi) is widely used to suppress gene expression in a specific manner. The efficacy of RNAi is mainly dependent on the sequence of small interfering RNA (siRNA) in relation to the target mRNA. Although several algorithms have been developed for the design of siRNA, it is still difficult to choose a really effective siRNA from among multiple candidates. In this article, we report the development of an image-based, quantitative, ratiometric fluorescence reporter assay to evaluate the efficacy of RNAi at the single-cell level. Two fluorescence reporter constructs are used. One expresses the candidate small hairpin RNA (shRNA) together with an enhanced green fluorescent protein (EGFP); the other expresses a 19-nt target sequence inserted into a cassette expressing a red fluorescent protein (either DsRed or mCherry). Effectiveness of the candidate shRNA is evaluated as the extent to which it knocks down expression of the red fluorescent protein. Thus, the red-to-green fluorescence intensity ratio (appropriately normalized to controls) is used as the read-out for quantifying the siRNA efficacy at the individual cell level. We tested this dual fluorescence assay and compared predictions to actual endogenous knockdown levels for three different genes (vimentin, lamin A/C and Arp3) and twenty different shRNAs. For each of the genes, our assay successfully predicted the target sequences for effective RNAi. To further facilitate testing of RNAi efficacy, we developed a negative selection marker (ccdB) method for construction of shRNA and red fluorescent reporter plasmids that allowed us to purify these plasmids directly from transformed bacteria without the need for colony selection and DNA sequencing verification.", "keywords": [ "RNA interference (RNAi) has become as an important tool not only for the identification of gene function but also for therapeutic applications1–3. RNAi is mediated by small interfering RNA (siRNA)", "typically 21-nt in length. The siRNA", "together with other cytoplasmic protein factors", "forms the RNA-induced silencing complex (RISC). The RISC recognizes the target mRNA through base pair matching", "and degrades the mRNA by cleavage of the siRNA/mRNA base-matching region. As a commonly used method", "a small hairpin RNA (shRNA) is experimentally expressed in cells", "after which an endogenous activity (DICER) removes the loop of the hairpin to generate a functional siRNA. As a result", "the expression of genes of interest can be specifically silenced1." ], "content": "Introduction\n\nRNA interference (RNAi) has become as an important tool not only for the identification of gene function but also for therapeutic applications1–3. RNAi is mediated by small interfering RNA (siRNA), typically 21-nt in length. The siRNA, together with other cytoplasmic protein factors, forms the RNA-induced silencing complex (RISC). The RISC recognizes the target mRNA through base pair matching, and degrades the mRNA by cleavage of the siRNA/mRNA base-matching region. As a commonly used method, a small hairpin RNA (shRNA) is experimentally expressed in cells, after which an endogenous activity (DICER) removes the loop of the hairpin to generate a functional siRNA. As a result, the expression of genes of interest can be specifically silenced1.\n\nOne of the biggest technical obstacles in using RNAi techniques is that the efficacy of gene silencing varies among different siRNA molecules4–6. The difference in efficacy of siRNA molecules is mainly dependent on the first 19-nt sequence of the sense strand of siRNA. Based on systematic analyses of the common features of highly effective siRNAs, several theoretical algorithms have been developed for the design of highly effective siRNA against mRNA targets7–9. However, the designed sequences often fail to behave as predicted and, therefore, it is still necessary to evaluate the efficacy of candidate siRNAs by experimental methods. Standard methods of evaluating RNAi have been Western blotting, immunofluorescence and quantitative RT-PCR. Alternative experimental approaches using reporter assays based on luciferases or fluorescent proteins have also been developed10–12. In these reporter assays, a DNA fragment of the gene to be silenced is either inserted into the untranslated region (UTR) of a reporter gene, or connected to the reporter in frame to express a fusion protein. After co-transfection of an siRNA or shRNA expression plasmid with the reporter plasmid, the enzymatic activity of luciferases or the fluorescence intensity of fluorescent proteins is measured. Although both standard and reporter methods have been useful in evaluating RNAi efficacy, a significant limitation is that they are generally applied to populations of cells. If expression of siRNA is not uniform in the population or if only a subset of cells expresses the knockdown construct, population methods of evaluating knockdown will introduce ambiguity into the results. Such ambiguity could be removed if the effectiveness of siRNA could be evaluated in individual cells.\n\nIn this article, we report the development of a quantitative, ratiometric reporter assay at the single-cell level. This assay is based on two technical advancements. The first is an improvement in the fluorescence reporter assay itself based on the use of quantitative fluorescence microscopy. Two fluorescence reporter constructs are used. One expresses the candidate shRNA together with a green fluorescent protein (EGFP); the other expresses a 19-nt target sequence inserted into a cassette expressing a red fluorescent protein (either DsRed or mCherry). Effectiveness of the candidate shRNA is evaluated depending on the extent to which it knocks down expression of the red fluorescent protein. Thus, the red-to-green fluorescence intensity ratio (appropriately normalized to controls) is used as the read-out for quantifying siRNA efficacy at the individual cell level. The image acquisition and analysis in our assay is simple and straightforward as only a standard fluorescence microscope system and image analysis software are necessary. The second advance is an improvement in the methods of preparing the required constructs, both the shRNA expression and reporter plasmids. By using a negative selection marker in the transforming plasmid, the ccdB gene, virtually all colonies of transformed bacteria contained the correct plasmid constructs. As a result, shRNA expression and red fluorescence reporter plasmids could be prepared without the need for colony selection and DNA sequence verification. We evaluated our dual fluorescence reporter assay using three test genes, vimentin, lamin A/C and Arp3, and confirmed that our fluorescent protein-based reporter assay could successfully predict siRNA efficacy with high fidelity. Taken together, these technical improvements facilitate the selection of highly effective siRNA among multiple candidates and make the ratiometric, fluorescence reporter assay practical and useful at the single-cell level.\n\n\nResults\n\nThe basic design concept of the assay is to express the candidate shRNA in one construct (the knockdown construct) that also expresses a reference (green) fluorescent protein and to express the target sequence in another construct (the target construct) that also expresses a reporter (red) fluorescent protein. The target sequence is identical to a sequence in the mRNA intended to be knocked down and is thus a proxy for the mRNA. The two types of construct were designed to be as similar as possible so that transfection and expression levels would also be similar.\n\nIn the target construct (denoted pREFLECT), the 19-nt target sequence is inserted into the 3′ UTR of the reporter DsRed2 or mCherry gene (Figure 1A, Supplemental Figure 1B). Candidate shRNAs are inserted into the knockdown construct, pSHIN-G expressing EGFP. The experiment consists of co-transfecting the knockdown construct with the target construct and comparing the results to co-transfection of the target construct with the empty pSHIN-G vector. When siRNA works effectively, the reporter mRNA will be degraded by RNAi via the target sequence in the 3′-UTR, resulting in reduction of red fluorescence.\n\n(A) Experimental design. Two plasmids (pSHIN-G and pREFLECT-R) are used. The plasmid pSHIN-G expresses shRNA and GFP simultaneously, whereas pREFLECT-R expresses a red fluorescent protein (RFP), DsRed2 or mCherry. The 3′-UTR of the RFP mRNA contains a 19-nt sequence target of siRNA to be tested. If siRNA, which is generated from shRNA, is effective, RFP expression is suppressed, resulting in reduction of red fluorescence. (B) The map of pREFLECT-R. RFP expression is driven under the SRα promoter. A 19-nt target sequence is inserted between NotI and XbaI sites. (C) Immunofluorescence of HeLa cells transfected with pSHIN-G-Vim-T4 and -T5. Phase contrast, vimentin-staining and EGFP images are shown. Arrowheads in the phase contrast images indicate transfected cells as shown by EGFP expression. The Vim-T5 shRNA expression suppressed vimentin significantly, while Vim-T4 did not show the RNAi effect. Bar, 20 µm. (D) Fluorescence images of co-transfection of pSHIN-G and pREFLECT-R(DsR) derivatives. Rat2 cells were transfected with pSHIN-G-Vim-T4 plus pREFLECT-R(DsR)-Vim-T4 (Vim-T4; +shRNA), pSHIN-G empty plasmid vector plus pREFLECT-R(DsR)-Vim-T4 (Vim-T4; -shRNA), pSHIN-G-Vim-T5 plus pREFLECT-R(DsR)-Vim-T5 (Vim-T5; +shRNA) or pSHIN-G empty plasmid vector plus pREFLECT-R(DsR)-Vim-T5 (Vim-T5; -shRNA). Two days after transfection, cells were fixed and DNA-stained with Hoechst33342. DNA, EGFP and DsRed2 fluorescence images are shown. The red fluorescence was not changed by shRNA expression of the Vim-T4 shRNA, whereas the Vim-T5 shRNA reduced red fluorescence significantly. Bar, 20 µm.\n\nThe initial evaluation of the dual construct system was carried out on the human vimentin gene. We chose six siRNA target sequences (Vim-T1 to -T6; Table 1), previously reported by other groups5,8. For each target sequence, the shRNA expression plasmids (pSHIN-G-Vim-T1 to -T6) were constructed as reported in the literature13. Five days after transfection of HeLa cells, immunofluorescence revealed that reduction of vimentin differed dependent on the target sequences. Vim-T4 did not show detectable reduction of vimentin (Figure 1C), whereas reduction was observed for the others, albeit at differing levels (Supplemental Figure 2). In particular, with Vim-T5, some cells lacked vimentin almost completely (arrowed cells in Figure 1C). The overall results are in good agreement with the previous reports5,8.\n\nHaving confirmed the relative activity of six different shRNAs against endogenous vimentin, we then tested whether the cognate 19-nt target sequences would show the same relative activity. Red fluorescence reporter plasmids were constructed by insertion of annealed oligonucleotides to the 3′-UTR of the DsRed2 red fluorescence protein (pREFLECT-R(DsR)-Vim-T1 to -T6; Figure 1B). Then the reporter plasmid was co-transfected with the corresponding shRNA expression plasmid (test experiment) or the empty pSHIN-G plasmid vector without hairpin sequence (control experiment) at the molar ratio of 4:1. The rat fibroblastic line, Rat2, was used for this assay because the Rat2 cells had nuclei with homogenous morphology, and this morphological feature made image analysis easy (see below). Two days after transfection, the red fluorescence levels of expressed DsRed2 were compared between the test and control samples. As shown in Figure 1D, Vim-T5 reduced red fluorescence remarkably, while for Vm-T4 DsRed2 expression was similar between the test and control samples. The suppression of red fluorescence agreed with the silencing activities of siRNA for endogenous vimentin (Figure 1D & Supplemental Figure 3). Thus, a simple 19-nt sequence in the context of our fluorescence reporter assay seemed to work, at least qualitatively, as a proxy for mRNA knockdown.\n\nNext, we developed a microscopy-based quantification method for the dual fluorescence reporter assay (Figure 2A). Transfected Rat2 cells were stained with the Hoechst33342 DNA dye to image and define nuclear regions. Although the expressed fluorescence was distributed throughout the cell, quantification of the green and red intensities was easier for the nuclear regions. DNA staining allowed us to define clear-cut regions for fluorescence measurement at individual cell level with assistance of image analysis software. The intensity of green and red fluorescence was recorded for each defined region. The data relative to transfectants were segregated from the data relative to untransfected cells by setting a threshold on the green fluorescence intensity. In control cells, as shown in Figure 2B (& Supplemental Figure 4A), green and red fluorescence intensities were linearly correlated over a wide range of expression levels. This result indicates that red (R)/green (G) ratios at the individual cell level can be used for statistical analysis. The observed linearity was likely related to the similarity of the expression cassettes between pSHIN-G and pREFLECT-R (Supplemental Figure 1), which use the identical promoter and polyA additional signal, as we could not observe good correlation between green and red fluorescence when EGFP and DsRed2 were expressed under different promoters. The length of the 3′-UTR was also likely important. When the full length of human vimentin coding sequence was inserted into pREFLECT-R(DsR), the expression level of red fluorescent protein was reduced significantly as compared to the 19-nt sequence and the green and red fluorescence intensities did not correlate as well (Supplemental Figure 5).\n\n(A) Scheme of image analysis. Interphase nuclei are defined from DNA-staining images. Although the expressed fluorescence is distributed throughout the cell, regions of interest defined by the DNA staining are convenient for measurement of the GFP and RFP intensities of individual cells. The fluorescence intensities were thresholded, averaged over the region of interest and subjected to statistical analysis (see Materials and Methods in detail). (B) Scatter plot of green and red intensities. Rat2 cells were co-transfected with pSHIN-G empty plus pREFLECT-R(DsR)-Vim-T5. Linear correlation between green and red fluorescence was observed over a wide range of expression levels. (C) Histograms of the ratios of red intensity to the green intensity (R/G). Left, co-transfection of pSHIN-G-Vim-T5 plus pREFLECT-R(DsR)-Vim-T5; right, co-transfection of pSHIN-G empty plus pREFLECT-R(DsR)-Vim-T5. The right histogram was created from the same data set as for Figure 2B. Upon shRNA expression (left histogram), the R/G ratios were significantly reduced because of RNAi of RFP. The medians, averages and standard deviations (Std) of the R/G ratios, as well as the numbers of analyzed cells (n), are shown in the insets.\n\nIn the test experiments using pSHIN-G plasmids with high RNAi efficiency such as Vim-T3 and -T5, the R/G ratios were much lower than those of the control experiments (Figure 2C and Supplemental Figure 4B&C). As shown Figure 2C, the distribution of the R/G ratios was different from the Gaussian distribution in the test experiments, indicating the median, rather than the average, more suitably represented the population of R/G ratios. In the test experiments, a small number of outliers with unusually bright red fluorescence were constantly observed (arrows in Supplemental Figure 4B), which affected the average of the R/G ratios as compared to their median (Figure 3C). The outliers imply that RNAi worked much less effectively in a few cells. Interestingly, such RNAi-insensitive cells were also observed in the case of knockdown of endogenous genes13,14.\n\nThe KD indexes were calculated from the medians of the R/G ratios. Bar graphs show the averages and standard deviations of the KD indexes in two or three experiments for each test system. (A, B) vimentin; (C, D) lamin; (E, F), Arp3. (A) KD indexes for Vim-T1 to -T6. The KD indexes of Vim-T2, -T3, -T5 and -T6 are also shown in the inset on an expanded scale. (B) Western blotting of HeLa cells for vimentin. HeLa cells were electroporated with each of the pSHIN-G-Vim constructs. In each lane, 30 µg of the protein sample was loaded. After detection with anti-vimentin antibody, the same membrane was analyzed by anti-α-tubulin antibody (loading control). (C) KD indexes for Lmna-T1 to -T8. (D) Quantification of expression level of laminA/C. The laminA/C expression levels of transfectants as assayed by immunofluorescence were normalized to the averages of untransfected cells. In each experiment, the median of the laminA/C of 50–225 transfected cells was calculated. The averages and standard deviation of two experiments are shown. (E) KD indexes for Arp3-T1 to -T6. The graph in the inset uses an expanded scale for the y-axis. (F) Western blotting of B16S cells for Arp3. B16S cells were transfected with one of pSHIN-puro-Arp3 constructs, and selected in the presence of puromycin for 2 days. For transfected and puromycin-selected cells (Arp3-T1 to -T6, and neg), 18 µg of the protein sample was loaded per lane. For untransfected cells (WT), a dilution series of the protein sample was loaded. The same membrane was analyzed by anti-Arp3 and anti-α-tubulin antibodies (loading control).\n\nWe defined the knockdown index (KD index) using the medians of the R/G ratios in the test and control experiments (mtest and mcontrol, respectively);\n\nKD = 100 * mtest/mcontrol * δ\n\nwhere δ is the correction coefficient introduced to adjust differences in EGFP expression levels between the shRNA expression plasmid and the empty vector. The correction coefficient (δ) was experimentally determined (see Supplementary materials for more detailed explanation) by using pREFLECT-R with a 19-nt scrambled sequence. In most cases, δ was nearly 1, implying that the EGFP expression level was not changed by the insertion of the hairpin sequence under human H1 promoter in pSHIN-G. However, for some target sequences, we found that EGFP expression was reduced by the insertion of the hairpin sequence. The KD index is a measure of the extent to which the target remains after knockdown by the candidate shRNA relative to empty vector control and is expressed in per cent. Thus, a KD of 0% signifies the strongest knockdown (no remaining target) whereas 100% signifies no knockdown.\n\nThe KD indexes of Vim-T1 to -T6 were calculated for acquired images (Figure 3A). Silencing of endogenous vimentin was analyzed by Western blotting of HeLa cell extracts 5 days after transfection of pSHIN-G-Vim plasmids (Figure 3B). The KD indexes were correlated to the gene silencing of endogenous vimentin. Among the six tested shRNA expression plasmids, Vim-T5 had the lowest KD index and showed the greatest knockdown to endogenous vimentin in Western blotting (Figure 3B).\n\nFor our assay system to be generally useful, it should work on any potential target. To test the generality of the assay, we evaluated silencing of two additional targets, lamin A/C and Arp3, a component of the Arp2/3 complex, as below. For silencing of human lamin A/C, we selected eight new target sequences (Table 1). Four target sequences (Lmna-T1 to -T4) were chosen simply according to the Tuschl’s criteria15, whereas the computer algorithms provided by Invitrogen (http://rnaidesigner.invitrogen.com/rnaiexpress) and Dharmacon (http://dharmacon.com/sidesign/) were used to select Lmna-T5 to -T8 targets. The KD indexes were calculated as shown in Figure 3C. The knockdown levels of endogenous lamin A/C were estimated by analyzing immunofluorescence of HeLa cells 5 days after transfection of pSHIN-G-Lmna plasmids (Figure 3D). Our previous work showed that analysis of immunofluorescence gave more reproducible and quantitative results about RNAi of laminA/C than Western blotting13. The results of the ratiometry (Figure 3C) were generally in good agreement with RNAi of endogenous lamin A/C. Seven of the eight tested shRNAs behaved as predicted by dual fluorescence assay. The KD indexes indicated the Lmna-T1 (1.4%) and -T8 (2.8%) as the best two sequences for RNAi. In the knockdown experiments of lamin A/C in HeLa cells, Lmna-T1 and -T8 showed the lowest remaining level of lamin A/C (12.7% and 11.9%, respectively). Only Lmna-T6 showed discordant behavior. The Lmna-T6 shRNA expression gave a low KD index of 4.3% but did not effectively silence endogenous lamin A/C (Figure 3C). Possibly the presence of a local secondary structure of endogenous lamin A/C mRNA might have restricted the access of the siRNA resulting in poor silencing, as previously seen in other cases11.\n\nIn a third test system, we applied our fluorescence reporter assay to select highly effectively siRNA to Arp3. From the common sequence between mouse and rat Arp3 genes, we picked up six target sequences (Arp3-T1 to -T6; Table 1). We carried out the dual fluorescence reporter assay and the ratiometry image analysis. The resulting KD indexes predicted that five siRNA sequences (Arp3-T1, Arp3-T3 to -T6) would be efficient (Figure 3E). To confirm endogenous Arp3 knockdown, we constructed shRNA expression plasmids with a puromycin resistance marker (pSHIN-puro-Arp3-T1 to -T6; Supplemental Figure 6A). One day after transfection of pSHIN-puro-Arp3 plasmids to mouse melanoma B16S cells, transfectants were selected in the presence of puromycin for 54 hours. The cell extracts were prepared 5 days after transfection and analyzed by Western blotting. In agreement with the KD indexes, five shRNA expression plasmids except Arp3-T2 reduced endogenous Arp3 proteins to 10–20% level of control cells (Figure 3F & Supplemental Figure 6B). The best result was obtained with Arp3-T5, which had the second lowest KD index value. As the Arp3-T1 and -T4 sequences are shared with human Arp3 as well, we further examined gene silencing by Arp3-T1 and -T4 in human SCC9 carcinoma cells (Supplemental Figure 6C). Both constructs silenced Arp3 gene expression significantly as shown in Supplemental Figure 6C for human SCC9 cells.\n\nIn summary, our fluorescence reporter assay successfully identified highly effective siRNAs for all three test cases and is likely, therefore, to be generally useful.\n\nThe above results indicate that our dual fluorescence assay is useful to estimate the RNAi efficacy of siRNA. However, the assay requires the construction of appropriate pairs of reporter plasmids which is labor intensive and might limit the general use of this reporter assay. In order to make the dual fluorescence assay more convenient, we modified the procedures of plasmid construction so as to speed up the workflow.\n\nThe key feature of the modification was to incorporate a negative bacterial selection marker. Conventionally, both shRNA expression and fluorescence reporter plasmids are constructed by insertion of annealed oligonucleotides. To obtain the proper plasmid constructs more efficiently, a negative bacterial selection marker, the ccdB gene16, was added at the insertion site of oligonucleotides (Supplemental Figure 7A&B). The ccdB gene prohibits these parental vectors (pSHIN-G[ccdB] and pREFLECT-R[ccdB]) to propagate by normal Escherichia coli strains. When annealed oligonucleotides are inserted in these vectors to remove the ccdB gene, these pSHIN-G and pREFLECT-R derivatives then allow bacterial growth. As a result, virtually all colonies are expected to contain the correct plasmid. Additionally, we used BstXI sites for the oligonucleotide insertion points. As shown in Supplemental Figure 7C&DBstXI-digested overhangs allow unidirectional cloning and prevent insertion of tandems of oligonucleotide duplexes. We tested the success rates of plasmid construction with these modified plasmid vectors. In the case of pREFLECT-R[ccdB], all plasmids contained the correct insert (10 out of 10 colonies). For pSHIN-G[ccdB], we used four short oligonucleotides instead of two long oligonucleotides for plasmid construction as shown in Supplemental Figure 7C. The sense and antisense parts were separately annealed and ligated in tandem to the plasmid vector. Although tandem ligation is thought to be unfavorable, DNA sequencing of purified plasmids revealed that more than 90% of bacterial colonies contained the correct product (13 out of 14 clones). One plasmid contained a 1-nt deletion in the insertion, which was possibly derived from a mistake during DNA oligonucleotide synthesis.\n\nConventionally, DNA clones need to be purified from several bacterial colonies and verified by DNA sequencing (left flow in Figure 4A). However, based on the above results, we hypothesized that plasmid DNA directly purified from transformed E. coli would work as well as a construct verified by DNA sequencing (the right flow in Figure 4A). We tested this idea using pSHIN-G-Lmna-T1 and -T4 shRNA expression plasmids that were directly purified from transformed E. coli without colony selection. As shown in Figure 4B, lamin A/C expression was efficiently suppressed by these plasmid constructs. We quantified reduction of lamin A/C at the individual cell level (Figure 4C and Supplemental Figure 8). The silencing was similar for Lmna-T1 and -T4 using both the direct purification method and the conventional plasmid preparation. These results indicate that, for test experiments, the shRNA expression plasmid can be prepared by using our new vector without colony selection.\n\n(A) The left flow indicates the conventional method of expression plasmid preparation including colony-selection and DNA sequencing. The right flow indicates the novel method using plasmid vectors with the negative selection marker (ccdB). The ccdB gene allows growth of bacteria containing plasmid with insertion of oligonucleotides, but not growth of the parent plasmid vectors. As a result, plasmid DNA can be purified from heterogeneous bacterial culture after transformation without colony-selection and DNA sequencing. (B) Immunofluorescence of lamin A/C in HeLa cells. HeLa cells were transfected with pSHIN-G-Lmna-T1 or -T4 prepared by the “no selection” method (the right flow in A). DNA-staining, EGFP and immunofluorescence of lamin A/C are shown. Reduction of lamin staining for cell expressing EGFP indicates knockdown. Bar, 20 µm. (C) Box plots of the lamin A/C levels [%] of HeLa cells transfected with shRNA expression plasmids. The top, bottom, and line through the middle of the box correspond to the 75th percentile (top quartile), 25th percentile (bottom quartile), and 50th percentile (median) respectively. The whiskers indicate the 10th percentile and 90th percentiles. The closed circle represents the mean. Two different methods were used for DNA preparation. The “selected” plasmid DNA was prepared by the conventional method (the left flow in A). Preparation of the “no selection” plasmid omits colony-selection and DNA-sequencing as shown in the right flow in A. The “no selection” procedure produced results comparable to the conventional procedure. The relative lamin A/C level was calculated as in Figure 3D. In each sample, 129–146 EGFP-expressing cells were analyzed.\n\nThe original experimental design of the dual fluorescence assay (Figure 1A) is theoretically ideal as a correlation between EGFP and shRNA is guaranteed by the expression of both the reporter and the hairpin RNA from the same plasmid. However, a drawback of this approach is the need to construct each fluorescent reporter-shRNA plasmid individually. Again, in the interest of speeding workflow and moving towards high throughput, we sought to overcome this drawback using a simplified three-plasmid approach (Figure 5A). In this approach, the EGFP reporter and the shRNA are expressed on different plasmids with the target carried on a third plasmid as before. With this approach, the EGFP reporter plasmid can be made essentially identical to the target red fluorescent reporter plasmid and the shRNA plasmids can be used without further modification. We tested this system on silencing human vimentin. We used mCherry as the red fluorescence reporter, instead of DsRed2, since mCherry maturates faster than that of DsRed217. The red fluorescence became sufficiently bright just 24 hours after transfection of mCherry, whereas DsRed2 required 48 hours. The pREFLECT-G plasmid with a scrambled 19-nt target sequence was constructed by replacing the DsRed2 part of pREFLECT-R(DsR) control plasmid with EGFP. Transfection of the EGFP control and mCherry reporter plasmids to Rat2 cells showed good linear correlation between the green and red fluorescence intensities (Figure 5B).\n\n(A) We used three plasmids: the shRNA expression plasmid, the pREFLECT-G and the pREFLECT-R. The pREFLECT-G plasmid expresses GFP as a reference, whereas the pREFLECT-R plasmid expresses RFP as a reporter for RNAi. The 3′-UTRs of the GFP and RFP mRNA contain 19-nt scrambled and target sequences, respectively. (B) Scatter plots of green and red fluorescence intensities. Rat2 cells were transfected with pREFLECT-G-Ctrl-T1 and pREFLECT-R(mCh)-Vim-T3 with or without pSHIN-Vim-T3. The cells were fixed 24 hr afterwards, and the green and red fluorescence intensities were measured. Whereas the green and red fluorescence intensities were both strong and showed linear correlation without shRNA expression (blue dots), the red fluorescence was significantly suppressed by shRNA expression (red dots). (C&D) Histograms of the R/G ratios. The results of co-transfection of pREFLECT-G-Ctrl-T1, pREFLECT-R(mCh)-Vim-T3 with and without pSHIN-Vim-T3 are shown (the left and right histograms, respectively). In C, pREFLECT-R(mCh)-Vim-T3 and pSHIN-Vim-T3 were prepared by the conventional method (the left flow in Figure 4A). The no-selection method (the right flow in Figure 4A) was used for both plasmids in D. Note expanded scale on left histograms demonstrating efficacy of the 3-plasmids procedure. (E) Knockdown (KD) indexes of Vim-T1 to -T6 target sequences. The bar graph shows the averages and standard deviation of the KD indexes in two independent experiments. Plasmids were prepared by the no-selection method. The results were similar to those obtained by the conventional plasmid procedure as shown in Figure 3. The KD indexes of Vim-T2, -T3, -T5 and -T6 are also shown in the inset.\n\nAs expected, when a shRNA expression plasmid with strong silencing efficacy (Vim-T3) was co-transfected, the red fluorescence intensity level was reduced significantly compared to the green fluorescence intensity level (Figure 5B). The outliers with unusually bright red fluorescence were not increased significantly compared to the two plasmid assay, implying that three plasmid co-transfection worked efficiently. Next, we tested whether the above simplified plasmid preparation without colony selection was applicable to the dual fluorescence assay using the three plasmids. The plasmid vectors containing the ccdB gene (pREFLECT-R(mCh)[ccdB] and pSHIN[ccdB]) were used to prepare mCherry reporter and shRNA expression plasmids, respectively. The distributions of the R/G ratios in the presence or absence of shRNA expression plasmid were similar between the conventional and simplified plasmid preparation (Figure 5C and D).\n\nThe KD indexes were calculated similarly to the original ratiometry assay, using the medians of the R/G ratios (mtest and mcontrol for the samples plus and minus the shRNA expression plasmid respectively). In the case of the three plasmid assay, the KD index is simply defined as mtest/mcontrol without the correction factor. We calculated the KD indexes for six target sequences, Vim-T1 to -T6. The results (Figure 5E) were similar to those obtained by our original fluorescence reporter assay using two plasmids (Figure 3A). Therefore the dual fluorescence assay using three plasmids worked as well as the originally developed assay using two plasmids.\n\n\nDiscussion\n\nWe have developed a ratiometric fluorescence reporter assay for siRNA efficacy based on image analysis at the individual cell level. A key advantage of an assay at the single-cell level is that it obviates ambiguities attendant upon methods that average results over populations of cells. With this assay we tested the interaction between various siRNAs and their target sequences in human, mouse and rat cells under the same experimental conditions. Therefore, the results can indicate the intrinsic gene silencing efficacies of different siRNA candidates. Our approach was validated by testing 20 different shRNAs for three different protein systems and, for all shRNAs except one, the estimation by the reporter assay was in good agreement with actual endogenous protein knockdown. The one exception was Lmna-T6. The main difference between our reporter assay and endogenous gene silencing is the surrounding RNA sequences present in the endogenous system. Therefore, we speculate that presence of local secondary structures of lamin A/C mRNA or RNA-binding proteins near the Lmna-T6 sequence might have restricted target recognition by siRNA. It is noteworthy that some candidate sequences (Lmna-T5 and -T7), which were selected by computer algorithms, did not show highly effective gene silencing both in our reporter assay and to endogenous protein. In a series of experiments for three other genes, we first selected five candidate sequences by computer algorithms, but only 1–3 sequences had the KD indexes less than 5% in each case, suggesting that not all candidate siRNAs would have sufficiently strong RNAi efficacy (unpublished data). These results indicate limitations in current theoretical prediction methods and suggest that a realistic RNAi effort still requires testing multiple siRNA targets to validate those that are effective.\n\nOur dual fluorescence assay has the advantage of speeding up and making a more reliable identification of effective siRNAs. We achieved results in only 24 hours, whereas endogenous protein reduction by RNAi usually requires more than 48 hours. In the case of endogenous gene silencing, pre-synthesized proteins are still present when siRNA eliminates the mRNA completely. As a result, it takes a long time before the effects appear at the protein level. The necessary time depends on the protein turn-over rate, but usually 24 hours are not sufficient to reduce the level of endogenous proteins4. In contrast, the synthesis of shRNA and mRNA encoding red fluorescence protein starts simultaneously after transfection in our reporter assay. As a result, red reporter mRNA would be degraded by functional siRNA before red fluorescent protein is translated and accumulated. Thus our assay can predict the siRNA efficacy faster than endogenous gene knockdown experiments. Additionally, this is a good feature for selecting siRNAs whose silencing might be toxic, as the assay will be able to show effective siRNA before the hazardous knockdown phenotypes appear.\n\nWe also modified the originally developed two plasmid assay system to accommodate situations where researchers have already invested in existing shRNA expression plasmids. In this situation, the three plasmid system (red reporter, green internal control and existing shRNA) may be of particular utility. Another technical improvement is the simplified construction of shRNA expression plasmids that allow omitting colony selection and sequencing verification. It should be noted that all the steps of simplified plasmid construction, including synthesis and annealing of DNA oligonucleotide, bacterial culture and plasmid DNA purification, are compatible with robotic high throughput platforms. Since our assay is microscopic, automated microscopes for high content screening18,19 will likely be useful for image acquisition and analysis. Consequently, by this method, preparation of shRNA expression plasmids for high-throughput screening is facilitated. Chemically synthesized siRNA may also be evaluated.\n\nFinally, it should be noted that our microscopic dual reporter assay approach is not limited to siRNA, but could also be applied to other nucleic acid targeting systems. For example, an attractive and possible application of our dual fluorescence assay is the screening of microRNA (miRNA) target sequences. At present, hundreds of miRNAs have been discovered, but their biological functions are poorly understood20. Our assay could easily be applied to miRNA target identification by substituting shRNA expression with miRNA expression.\n\nRecently new genome editing procedures have been developed to modify gene expression. These include transcription activator-like effector nuclease (TALEN) and clustered regulatory interspaced short palindromic repeat (CRISPR)/Cas-based RNA-guided DNA endonucleases21. TALEN and CRISPR/Cas are based on DNA double strand breaks for targeting specific sequences in the genome. Since TALEN and CRISPR/Cas approaches work at the genomic level, they can produce permanent changes in the gene expression as opposed to RNAi which results in transient knockdown of expression. At present, however, several technical obstacles limit TALEN and CRISPR/Cas approaches from general use on culture cells. By both methods, targeting rates were not sufficiently high enough for mammalian culture cells to disrupt both copies of genes22–25. In addition, off-target genome disruption, as for RNAi, will need to be considered26–30. Therefore, RNAi is likely to remain useful as a practical technology for suppressing gene expression. In summary, we have developed an image-based, dual fluorescence ratiometric assay to evaluate the knockdown efficacy of siRNA. The assay is microscopic, quantitative and relatively fast. The red fluorescence reporter and shRNA expression plasmids can be directly purified from transformed E. coli with or without colony selection and verification by DNA sequencing. A variant three-plasmid procedure allows the use of existing shRNA plasmids.\n\n\nMaterials and methods\n\nAll reagents were purchased from Sigma-Aldrich, unless noting specifically.\n\nHeLa cells, a human cervical adenocarcinoma cell line, were obtained from the American Tissue Culture Collection (ATCC). HeLa cells were grown in Eagle’s Minimum Essential Medium (MEM) containing Earle’s salts, 2 mM l-glutamine, 0.1 mM non-essential amino acids and 1.0 mM sodium pyruvate, supplemented with 10% fetal bovine serum (FBS). B16S mouse melanoma, Rat2 rat fibroblast and SCC9 human carcinoma lines were provided by Drs. V. Gelfand (Northwestern Univ.), F. Gertler (MIT) and K. Green (Northwestern Univ.), respectively. B16S and Rat2 cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% FBS. SCC9 cells were maintained in Dulbecco’s Modified Eagle Medium: Ham’s Nutrient Mixture F-12 (DMEM/F12, 1:1) media supplemented with 10% fetal bovine serum. All the cell lines were cultured at 37°C with 5% CO2. Culture media and FBS were purchased from Gibco and Atlanta Biologicals, respectively.\n\nThe parental plasmid vectors, pSHIN-G[ccdB], pREFLECT-R(DsR)[ccdB] and pREFLECT-R(mCh)[ccdB], were constructed as described in detail in the Supplementary materials. These plasmids containing the ccdB gene were amplified in a special E. coli strain, DB3.1 (Invitrogen). A shRNA expression vector without fluorescent protein reporter, pSHIN[ccdB], was constructed by BamHI digestion and subsequent self-ligation of pSHIN-G[ccdB].\n\nKnockdown constructs except for a negative control were constructed by using sets of four oligonucleotides that were obtained from Integrated DNA Technologies (IDT) at the desalting grade. The construction scheme was illustrated in Supplemental Figure 7C. The oligonucleotides had the following configurations:\n\n#1-fwd: 5′-GATCCCCxxxGC-3′,\n\n#1-rev: 5′-ACAGGAAGCyyyGGG-3′\n\n#2-fwd: 5′-TTCCTGTCACyyyTTTT-3′\n\n#2-rev: 5′-xxxGTG-3′\n\nwhere xxx and yyy represent a selected 19-nt target and its complementary sequence, respectively. The loop sequence (GCTTCCTGTCAC) comes from the human miR23 gene. Oligonucleotides #1-fwd and #1-rev were annealed in 30 mM HEPES, 100 mM K-acetate and 2 mM Mg-acetate (pH7.0) by heating at 95°C for 4 min, incubating at 70°C for 10 min and gradually cooling down to 4°C to make oligoduplex #1. Similarly oligoduplex #2 was prepared. Then the two duplexes were mixed and phosphorylated by T4 polynucleotide kinase in the presence of 0.1 mM ATP at 37°C for 30 minutes, followed by the reaction termination at 70°C for 10 minutes. For a negative control shRNA expression plasmid, two oligonucleotides,\n\n5′-GATCCCCATGTACTGCGCGTGGAGACTTCAAGAGAGTCTCCACGCGCAGTACATTTTT-3′\n\n5′-ATGTACTGCGCGTGGAGACTCTCTTGAAGTCTCCACGCGCAGTACATGGG-3′,\n\nwere annealed and phosphorylated as above. The phosphorylated oligos were ligated with BglII-BstXI digested pSHIN-G[ccdB] or pSHIN[ccdB] to create shRNA expression plasmids with or without EGFP marker, respectively.\n\nReporter constructs were constructed by insertion of two oligonucleotides: 5′-xxxTTCG-3′ and 5′-yyyTTGC-3′, where xxx and yyy represent the sense and antisense sequences of a selected 19-nt target respectively (Supplemental Figure 7D). A pair of oligonucleotides were annealed as above and ligated to a BstXI gap of pREFELECT plasmids. Neither phosphorylation of oligonucleotides nor dephosphorylation of the plasmid vector was necessary. All the target sequences in this experiment are summarized in Table 1.\n\nKnockdown plasmids with puromycin resistant markers were constructed as below. The phosphoglycerate kinase I (PGK) promoter and puromycin resistant (puroR) gene sequences were PCR-amplified from pSilencer™ 5.1-H1 Retro (Ambion) with oligonucleotide primers, 5′-AAATCTAGATACCGGGTAGGGGAGGCGCT-3′ and 5′-AAAGCGGCCGCTCAGGCACCGGGCTTGCGG-3′. After digestion with XbaI and NotI, the amplified fragment was inserted to an XbaI-NotI gap of pSHIN-G to create pSHIN-puro, which does not contain retroviral sequences. The pSHIN-puro-Arp3 plasmids were constructed by replacement of the expression cassette of EGFP with the one of puroR.\n\nStable knockdown plasmids for human cells were constructed as below. A BspHI DNA fragment containing pUC ori sequence was excised from pBluescript KS(+) (from Dr. F. Hanaoka, Gakushuin Univ.), while pLZRS-linker (from Dr. K. Green, Northwestern Univ.) was digested with BspHI to obtain a DNA fragment containing the oriP and EBNA-1 sequences as well as the puromycin resistant gene. These fragments were ligated to make pEpiso-KS(+). The shRNA expression cassettes were excised by BamHI-HindIII digestion from pSHIN-G-Arp3-T1, -Arp3-T4 and -neg, and inserted to a BamHI-HindIII gap of pEpiso-KS(+) to create pEpiso-shR[Arp3-T1], [Arp3-T4] and [neg].\n\nSmall scale plasmid DNA was purified from 1.5–2 mL bacterial culture with QIAprep Spin Miniprep kit (QIAGEN) or Perfectprep Plasmid Mini kit (Eppendorf). When using the Perfectprep Plasmid Mini kit, the TritonX-114 extraction step was added for removal of endotoxin. Briefly, after centrifugation of the alkali lysis extract, 3 µL of TritonX-114 was added to ~300 µL of the supernatant. The mixture was cooled on ice for 10 minutes, subsequently incubated at 55° C for 10 min and centrifuged for 5 minutes. The supernatant was used for the following DNA purification with 450 µL of the DNA binding matrix. By this modification, transfection efficiency to Rat2 cells was greatly improved. Competent cells of E. coli strains, TOP10 (Invitrogen) and XL10-Gold (Stratagene) with high competency (~109/µg) were used. Maxi-prep of plasmid DNA was carried out by using the EndoFree Plasmid Maxi kit (QIAGEN). All the clones were confirmed by DNA sequencing.\n\nRat2 cells, plated in 12 well plates, were co-transfected with a combination of plasmids by using the transfection reagent TransIT-LT1 (Mirus). For two-plasmid assays, 0.2 µg of a pSHIN-G derivative and 0.8 µg of a pREFLECT-R derivative were used. For three-plasmid assays, 0.2 µg of pSHIN derivative, 0.2 µg of pREFLECT-G-Ctrl and 0.8 µg of pREFLECT-R(mCh) derivative were transfected. After 3–4 hours of incubation with the plasmid DNA/transfection reagent mixture, the cell culture medium was replaced in order to minimize cytotoxicity by endotoxin. The cells were fixed with 4% formaldehyde for 30 minutes at the time point of 48 hour (for DsRed2) or 24 hour (for mCherry) after adding the plasmid DNA/transfection reagent mixture. The fixed samples were permeabilized with 1% TritonX-100 for 5 minutes and DNA-stained with 10 µg/mL Hoechst 33342 (Invitrogen) for 20 minutes. The samples were prepared either on glass coverslips or in plastic culture plates. Fluorescence microscopy was carried out using a DIAPHOTO 300 microscope (NIKON) equipped with a 10× dry objective lens (Plan 10, N.A.0.25, Ph1 DL; NIKON), a CCD camera (CH350; Photometrics) and a filter wheel system (LAMBDA 10-2; Sutter Instrument). For EGFP/DsRed2 imaging, the 86100bs quad filter (Chroma) was used with combinations of excitation/emission filters including D360/40× and S457/50m for DNA, S480/25× and S520/40m for EGFP, and S555/28× and S617/73m for DsRed2. For EGFP/mCherry imaging, the 51019bs EGFP/DsRed dual filter (Chroma) was used with combinations of D360/40× and S457/50m for DNA, S480/25× and S520/40m for EGFP, and S573/23× and S630/75m for mCherry. MetaMorph software (Molecular Devices) was used for image acquisition and analysis. For statistical analysis, Excel (Microsoft) and DeltaGraph (Redrock) software programs were used.\n\nImage analysis was as follow. From DNA-staining images, cellular nuclei were defined. The threshold value was determined manually as approximately 95% of DNA-stained regions had the fluorescence intensity over that value. The continuous regions with the fluorescence higher than the threshold value were recorded by MetaMorph software. Defined regions were ellipses corresponding to cellular nuclei. The defined regions were further filtered for the normal range of nuclear sizes (1.1 µm2–2.7 µm2) and shapes of Rat2 cells and served as regions of interest for further analysis. The average intensities in green (EGFP) and red (DsRed2 or mCherry) channels were recorded for each defined region of interest. The green and red intensities were calculated by subtraction of the backgrounds that were estimated using an untransfected sample of Rat2 cells. The transfectants were selected by thresholding the EGFP intensity. The threshold value was preset from background analysis of pilot experiments under identical imaging condition (see Supplementary materials). Cellular nuclei with EGFP levels above the threshold level were used for further analysis. The ratio of red to green signal (R/G) was calculated for each transfectant. Usually more than 300 data points were collected for each sample, and used to calculate the median. The knockdown index (KD) was calculated by comparison of the median R/G values as described in the Results section and the Supplementary materials.\n\nHeLa cells were transfected with pSHIN-G-Lmna or -Vim plasmids using the TransIT-LT1 reagent (Mirus) according to the manufacturer’s instruction. Five or six days after adding the plasmid/transfection reagent mixture, the cells were fixed with 4% formaldehyde for 30 minutes and permeabilized with 1% TritonX-100 for 5 minutes. Mouse monoclonal anti-laminA/C (clone 636, Santa Cruz Biotechnology, #sc-7292) or anti-vimentin (clone V9, Chemicon, #MAB3400) antibodies were added to the cells at 1:100 for 30 minutes. After washing with phosphate buffer saline (PBS), the samples were incubated with 10 µg/mL tetramethylrhodamine-conjugated donkey anti-mouse IgG (Jackson Immuno Research, #715-025-150) and 10 µg/mL Hoechst33258 (Invitrogen) for 20 minutes. Images were taken by the above microscope setting except for using 20× objective lens (Fluor 20, N.A.0.75, Ph3DL; NIKON). Endogenous lamin A/C was quantified as previously described13. Briefly, the red fluorescence of lamin A/C-staining was measured and normalized to the average intensity of untransfected cells that did not express EGFP. In each experiment, 50–225 EGFP-expressing cells were analyzed.\n\nFor knockdown of vimentin, HeLa cells were electroporated with pSHIN-G-vim plasmids as described elsewhere13. EGFP-expressing cells usually exceeded 75% of total cells. Five days after electroporation, the protein samples were prepared. For knockdown of Arp3 in B16S cells, pSHIN-puro-Arp3 plasmids were transfected with TransIT-LT1 (Mirus). One day after addition of the DNA/liposome mixture, the cells were cultured in DMEM with 10% FBS and 2.0 µg/mL puromycin for 9 hours, following additional 45 hours culture in the presence of 1.5 µg/mL puromycin. The selected B16S cells were cultured two additional days without puromycin. For knockdown of Arp3 in SCC9 cells, pEpiso-shR[Arp3-T1], [Arp3-T4] or [neg] were transfected using TransIT-LT1 (Mirus). One day after transfection, puromycin was added at the final concentration of 0.8 µg/mL. The transfectants were selected in the presence of 0.8 µg/mL of puromycin for 9 days. Preparation of protein samples, SDS-PAGE, and transferring to PROTRAN nitrocellulose membrane (Whatman) were described elsewhere13. Mouse monoclonal anti-vimentin (clone V9, Chemicon, #MAB3400, 1:1000), rabbit polyclonal anti-Arp3 (Upstate, #07-272 1:500) and mouse monoclonal anti-α-tubulin (clone B-5-1-2, SIGMA-ALDRICH, #T5168, 1:5000) antibodies were used. Horseradish peroxidase (HRP)-conjugated secondary antibodies were obtained from KPL and GE Healthcare. ECL was carried out using ECL Western Blotting Detection Reagent (GE Healthcare), SuperSignal West Pico Chemiluminescent Substrate (PIERCE) or Immobilon Western Chemiluminescent HRP Substrate (Millipore). The chemiluminescent signal was recorded by the LAS-3000 imager (Fuji Film) or detected using Hyperfilm ECL (GE Healthcare). The results were analyzed by ImageJ software.\n\n\nData availability\n\nZENODO: Data of quantitative, ratiometric dual fluorescence reporter assay at the single-cell level, doi: 10.5281/zenodo.829631", "appendix": "Author contributions\n\n\n\nSK and GB conceived the study. SK designed and conducted the experiments, and analyzed the results. SK and GB 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 research was supported by NIH grant GM 25062 to GGB and the Uehara Memorial Foundation Fellowship to SK.\n\n\nAcknowledgments\n\nWe thank Drs. F. Gertler (MIT), K. Green (Northwestern Univ.), V. Gelfand (Northwestern Univ.), R. Goldman (Northwestern Univ.), R. Tsien (UCSF) and F. Hanaoka (Gakushuin Univ.) for supplying the reagents. We also appreciate Drs. F. Hanaoka and I. Mabuchi (Gakushuin Univ.) for allowing us to use their experimental facility including the LAS-3000 imager system. We also thank Drs. V. Gelfand and D. Vignjevic (Institute Curie) for their critical reading and comments on the manuscript.\n\n\nSupplementary materials\n\n\n\n\nReferences\n\nBoutros M, Ahringer J: The art and design of genetic screens: RNA interference. Nat Rev Genet. 2008; 9(7): 554–566. PubMed Abstract \| Publisher Full Text\n\nCastanotto D, Rossi JJ: The promises and pitfalls of RNA-interference-based therapeutics. Nature. 2009; 457(7228): 426–433. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKetting RF: The many faces of RNAi. Dev Cell. 2011; 20(2): 148–161. PubMed Abstract \| Publisher Full Text\n\nElbashir SM, Harborth J, Lendeckel W, et al.: Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature. 2001; 411(6836): 494–498. PubMed Abstract \| Publisher Full Text\n\nHarborth J, Elbashir SM, Bechert K, et al.: Identification of essential genes in cultured mammalian cells using small interfering RNAs. J Cell Sci. 2001; 114(Pt 24): 4557–4565. PubMed Abstract\n\nMartinez J, Patkaniowska A, Urlaub H, et al.: Single-stranded antisense siRNAs guide target RNA cleavage in RNAi. Cell. 2002; 110(5): 563–574. PubMed Abstract \| Publisher Full Text\n\nReynolds A, Leake D, Boese Q, et al.: Rational siRNA design for RNA interference. Nat Biotechnol. 2004; 22(3): 326–330. PubMed Abstract \| Publisher Full Text\n\nUi-Tei K, Naito Y, Takahashi F, et al.: Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference. Nucleic Acids Res. 2004; 32(3): 936–948. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPei Y, Tuschl T: On the art of identifying effective and specific siRNAs. Nat Methods. 2006; 3(9): 670–676. PubMed Abstract \| Publisher Full Text\n\nTuschl T, Zamore PD, Lehmann R, et al.: Targeted mRNA degradation by double-stranded RNA in vitro. Genes Dev. 1999; 13(24): 3191–3197. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYoshinari K, Miyagishi M, Taira K: Effects on RNAi of the tight structure, sequence and position of the targeted region. Nucleic Acids Res. 2004; 32(2): 691–699. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHung CF, Lu KC, Cheng TL, et al.: A novel siRNA validation system for functional screening and identification of effective RNAi probes in mammalian cells. Biochem Biophys Res Commun. 2006; 346(3): 707–720. PubMed Abstract \| Publisher Full Text\n\nKojima S, Vignjevic D, Borisy GG: Improved silencing vector co-expressing GFP and small hairpin RNA. Biotechniques. 2004; 36(1): 74–79. PubMed Abstract\n\nFire A, Xu S, Montgomery MK, et al.: Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature. 1998; 391(6669): 806–811. PubMed Abstract \| Publisher Full Text\n\nElbashir SM, Lendeckel W, Tuschl T: RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 2001; 15(2): 188–200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGabant P, Szpirer CY, Couturier M, et al.: Direct selection cloning vectors adapted to the genetic analysis of gram-negative bacteria and their plasmids. Gene. 1998; 207(1): 87–92. PubMed Abstract \| Publisher Full Text\n\nShaner NC, Campbell RE, Steinbach PA, et al.: Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat Biotechnol. 2004; 22(12): 1567–1572. PubMed Abstract \| Publisher Full Text\n\nCarpenter AE: Image-based chemical screening. Nat Chem Biol. 2007; 3(8): 461–465. PubMed Abstract \| Publisher Full Text\n\nConrad C, Gerlich DW: Automated microscopy for high-content RNAi screening. J Cell Biol. 2010; 188(4): 453–461. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBartel DP: MicroRNAs: target recognition and regulatory functions. Cell. 2009; 136(2): 215–233. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaj T, Gersbach CA, Barbas CF 3rd: ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering. Trends Biotechnol. 2013; 31(7): 397–405. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller JC, Tan S, Qiao G, et al.: A TALE nuclease architecture for efficient genome editing. Nat Biotechnol. 2011; 29(2): 143–148. PubMed Abstract \| Publisher Full Text\n\nCong L, Ran FA, Cox D, et al.: Multiplex genome engineering using CRISPR/Cas systems. Science. 2013; 339(6121): 819–823. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMali P, Yang L, Esvelt KM, et al.: RNA-guided human genome engineering via Cas9. Science. 2013; 339(6121): 823–826. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakada S, Sato T, Ito Y, et al.: Targeted gene deletion of miRNAs in mice by TALEN system. PLoS One. 2013; 8(10): e76004. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoch J, Scholze H, Schornack S, et al.: Breaking the code of DNA binding specificity of TAL-type III effectors. Science. 2009; 326(5959): 1509–1512. PubMed Abstract \| Publisher Full Text\n\nFu Y, Foden JA, Khayter C, et al.: High-frequency off-target mutagenesis induced by CRISPR-Cas nucleases in human cells. Nat Biotechnol. 2013; 31(9): 822–826. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHsu PD, Scott DA, Weinstein JA, et al.: DNA targeting specificity of RNA-guided Cas9 nucleases. Nat Biotechnol. 2013; 31(9): 827–832. PubMed Abstract \| Publisher Full Text\n\nMali P, Aach J, Stranges PB, et al.: CAS9 transcriptional activators for target specificity screening and paired nickases for cooperative genome engineering. Nat Biotechnol. 2013; 31(9): 833–838. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPattanayak V, Lin S, Guilinger JP, et al.: High-throughput profiling of off-target DNA cleavage reveals RNA-programmed Cas9 nuclease specificity. Nat Biotechnol. 2013; 31(9): 839–843. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKojima S, Borisy GG: Data of quantitative, ratiometric dual fluorescence reporter assay at the single-cell level. ZENODO. 2014. Data Source" }	[ { "id": "3782", "date": "06 Mar 2014", "name": "Kenneth 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\nThis article introduces a new, fluorescence-based reporter assay for evaluating the knockdown efficacy of individual siRNA sequences. The authors effectively demonstrate the robustness of their assay in predicting effective siRNAs by direct comparison to endogenous protein knockdown, and they highlight the advantages this dual reporter assay provides over traditional methods, including significantly faster result times (24 hours versus 48+ hours) and evaluation of knockdown on the individual cell level. To expedite plasmid construction, the authors additionally modified their reporter plasmids with negative selection markers to bypass the need for DNA sequencing after addition of specific siRNA/target oligos. This very promising dual fluorescence reporter assay will be accessible to most standard cell and molecular biology laboratories, provides clear advantages over traditional methods of siRNA evaluation, and will provide avenues for high-throughput screening for effective siRNA sequences.Minor points:The rationale behind the three plasmid system is a bit confusing. While it appears to be designed to allow adaptability with previously constructed shRNA plasmids, it seems that the limitation of losing direct shRNA dosage correlation to GFP expression could sometimes be an offsetting disadvantage. The expression of shRNA could be significantly different than that of GFP, especially in hard-to-transfect cell lines, where it could potentially lead to false positives/negatives in the R/G ratios. Please clarify how the final bin of the R/G ratios are plotted, since the last bar in Figs. 2C, 5C, and 5D are all higher than adjacent bars. Does this last bin represent the sum of this particular X value and all higher values? Are these R/G final bins due to the “RNAi-insensitive cell” population mentioned on page 6? On page 7, the authors note that EGFP expression was reduced by insertion of the hairpin sequence for some target sequences. How severe was this reduction in the worst cases? In Fig. 3f, the tubulin control western blot staining does not seem to match the 10-fold dilution of WT at 1.8 compared to “neg” at 18 – which of the western blot kits was used, since it might not be linear? Very minor text issues should be corrected on page 8 (“since mCherry maturates faster than that of DsRed2”, page 11 (“unless noting specifically”), and page 13 (“Image analysis was as follow”).", "responses": [ { "c_id": "765", "date": "10 Apr 2014", "name": "Shin-ichiro Kojima", "role": "Author Response", "response": "Dear Drs. Kenneth Yamada and Matthew Kutys,Thank you very much for your comments.Comment #1: The potential risks/feasibility of the three plasmid system.Both reviewers pointed out the possible disadvantage of the dual fluorescence assay using three plasmids. They were concerned that the linkage between EGFP and shRNA expression is not guaranteed, which might create potential false positives and/or negatives. We agree with their comments and have modified the last line of the Results and added a discussion about this pitfall in accordance with their comments. However, the offsetting advantage is that the three-plasmid system makes possible the use of existing siRNA and shRNA libraries. Some researchers may find this advantage compelling. Therefore, we think that it is still worthwhile to report the three plasmid system. Comment #2: The histogramsThe last bin represents the sum of the values higher than the indicated maximum (0.12 in the case of the left histogram in Fig. 2C). To clarify the meaning of the first and last bins, we modified Fig. 2C and its legend. For other histograms, we expect that readers will be able to deduce the meaning of the first and last bins from Fig. 2C. As to the height of the last bins, as you might imagine, the RNAi insensitive outliers made the last bins much taller than the adjacent ones.Comment #3: Reduction of the EGFP intensity upon insertion of shRNA expression cassettes.For 4 out of 20 target sequences, the EGFP intensity was reduced by more than two fold (38%, 41%, 28% and 49% for Vim-T4, Lmna-T4, Lmna-T6 and Arp3-T4, respectively, compared to the control level).Comment #4: Western blottingFor blotting, we confirmed the protein amounts two different ways; Ponceau staining of the membranes and gel staining by GelCode Blue Stain Reagent (PIERCE). Thus, the result for a-tubulin is likely related to the insufficient linearity between the ECL signals and the protein amounts in Western blotting, as you pointed out. We used the SuperSignal West Dura substrate (PIERCE) for a-tubulin. We forgot to include the name of this substrate in the Materials and Methods section, so we have added it. All the ECL substrates used in this study gave reasonable linearity after optimization of experimental parameters (protein amounts, dilution of antibodies etc.). However, it is still difficult to get consistently good linearity over a wide range of protein amounts, since the ECL signals tend to saturate as the protein amount increases.Comment #5: Grammar mistakes.Thank you very much for reading through the manuscript very carefully. We have corrected grammar and typing mistakes in the revised manuscript." } ] }, { "id": "4230", "date": "07 Apr 2014", "name": "Wen Lu", "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 introduces a new dual fluorescence assay for quantitatively measuring the efficiency of siRNA-mediated mRNA knockdown in cultured cells. The authors used two fluorescent reporters to independently represent the levels of target gene expression and shRNA expression and then applied computational analysis to calculate the ratiometric fluorescence value between the two reporters as an index of siRNA efficiency. This new approach presents a precise measurement of siRNA knockdown efficiency by comparing the target gene expression level to shRNA expression level directly in single cells, and serves as a reliable method to test different shRNA sequences for siRNA knockdown. Overall it clearly provides at least three major advantages over the traditional methods (western blot, etc.) for siRNA knockdown evaluation:It measures the siRNA knockdown efficiency in individual cells so that it can bypass the potential problem of obtaining ambitious results from transfected cells that express the siRNA at various levels. Furthermore, they used automated image acquisition and analysis, which makes the approach feasible for high-throughput screening.\n\nIt significantly shortens the procedure of estimating siRNA efficiency. This method targets an exogenous reporter gene from a transfected vector, minimizing the concern of pre-synthesized endogenous proteins. Because of this, it takes less time to see siRNA effect after transfection (~24 hours) compared to the traditional methods (>48 hours).\n\nThe authors also introduced a negative bacteria selective marker, the ccdB gene, at the site of plasmid insertion, which may eliminate the colony selection step and significantly simplify the cloning procedure. Here are some points that I hope the authors could address:How can this method estimate the siRNA efficiency if the target sequences are embedded in a complicated secondary structure? The authors used 19-bp target sequence at the 3’UTR of a red fluorescent protein, and used the red fluorescent signal as a measurement of siRNA knockdown. In most cases this technique is sufficient, as the authors documented in this article. But there are some exceptions as the authors pointed out, for example, the Lmna-T6 target sequence (Figure 3C and 3D). I wonder whether the authors can incorporate more 5’ and 3’ target sequences into the pREFLECT vector to better test the efficiency at different local mRNA secondary structures. The advantage of a three plasmid system is not quite convincing. I think the true advantage of the two plasmid system is that the molar ratio of GFP and shRNA in each individual cell is fixed, and thus the GFP signal truly indicates how much shRNA is expressed in the same cell. By adopting the three plasmid system, this advantage is lost. The authors claimed that the three plasmid system can expedite the workflow because the GFP and mCherry vectors are essentially identical for all experiments, and the previously constructed shRNA vectors can be used without modification. Obviously, it could save some cloning time in constructing shRNA vector, but it creates issues in estimating shRNA expression level at the same time. For this reason, I do not see any significant advantage for testing siRNA targeting sequences using the three plasmid system over the two plasmid system.", "responses": [ { "c_id": "764", "date": "10 Apr 2014", "name": "Shin-ichiro Kojima", "role": "Author Response", "response": "Dear Dr. Wen Lu,Thank you for your valuable comments.Comment #1: Possibility of target sequences being embedded in complicated secondary structure.We agree with the reviewer that secondary structure in the target mRNA could render the prediction of shRNA efficacy invalid. Indeed, our paper documents such a case for a lamin mRNA and we note in the Discussion the potential complicating effect of secondary structure. However, of greater importance for the paper is that this was an exception - the method worked for almost all the targets tested. Theoretically, it is possible to examine the effects of the sequences adjacent to the 19-nt target by our dual fluorescence assay using reporter plasmids with inserted nucleotide sequences longer than 19 nt. Such plasmids could be constructed using synthesized oligonucleotides or PCR amplified fragments. Although such experimentation could provide new insights into mRNA silencing, we believe that this issue is really beyond the scope of the current paper.Comment #2: The feasibility of three plasmid system.See our response to Comment #1 of Drs. Kenneth Yamada and Matthew Kutys." } ] } ]	1	https://f1000research.com/articles/3-60
https://f1000research.com/articles/3-114/v1	16 May 14	{ "type": "Correspondence", "title": "Recent trends in airway management: we are not ready to give up fiberoptic endoscopy", "authors": [ "Davide Cattano", "Rabail Chaudhry", "Rashida Callender", "Peter V Killoran", "Carin A Hagberg", "Rabail Chaudhry", "Rashida Callender", "Peter V Killoran", "Carin A Hagberg" ], "abstract": "The purpose of this correspondence is to discuss recent findings related to current trends in airway management and to discuss the utilization rates of video laryngoscopes versus traditional techniques in USA, UK, and Canada. To highlight the increased use of video laryngoscopes in difficult airway situations, data on the use of alternative airway devices at our institution collected from 2008 to 2010 are presented alongside the results of previously published surveys collected from 2002 to 2013.", "keywords": [ "Education and research in anesthesia have increasingly focused on the management of difficult airways", "leading to the development of new devices that are gradually becoming available and part of routine use across the globe. It is rather interesting to assess whether we have made much progress in using such devices over the past decade." ], "content": "Correspondence\n\nEducation and research in anesthesia have increasingly focused on the management of difficult airways, leading to the development of new devices that are gradually becoming available and part of routine use across the globe. It is rather interesting to assess whether we have made much progress in using such devices over the past decade.\n\nWe read with great interest the letter ‘Should we really consider to lay down the Macintosh laryngoscope?’1, in which Merli G. et al. discuss the present and future roles of video laryngoscopes and the continued value of older instruments, i.e. the Macintosh direct laryngoscope. We agree with the authors that over the past two decades, a large number of airway devices have been introduced into clinical practice.\n\nData from the early 2000s suggest that, despite the widespread availability of newer airway equipment, traditional techniques (direct laryngoscopy, laryngeal mask airway (LMA), and flexible fiberoptic endoscopy) were the preferred techniques for intubation (Table 1). Ezri et al.2 reported in 2003 that US attending anesthesiologists preferably used flexible fiberoptic endoscopy (75%) for difficult airway management and preferred LMA (81%) in failed intubation/ventilation scenarios. Similarly, in 2004, fiberoptic endoscopy (64%) and some form of blind technique (26%) were used by anesthesiologists in the UK4. In 2005, practitioners in Canada preferred fiberoptic endoscopy (34%) and direct laryngoscopy (48%)5. In most surveys, lack of availability and training with newer equipment was of concern2–5.\n\nWe analyzed the utilization rates of alternative airway devices using data collected between 2008 and 2010 at our institution, the University of Texas Medical School at Houston, Memorial Hermann Hospital – Texas Medical Center (Table 2).\n\nThe most commonly used alternative airway devices were oral fiberoptic intubation (OFOI), (n=318, usage rate=3.69%, first attempt success rate=92.5%), the Glidescope® video laryngoscopy system (Verathon Inc, USA), (n=223, usage rate=2.59%, first attempt success rate=95.5%), the Storz C-MAC® video laryngoscopy system (Karl Storz, Germany), (n=154, usage rate=1.79%, first attempt success rate=94.8%), the Aintree Intubation Catheter (Cook Critical Care, USA), (n=106, usage rate=1.23%, first attempt success rate=96.2%), bougie (n=92, usage rate=1.07%, first attempt success rate=95.7%) and nasal fiberoptic intubation (NFOI), (n=92, usage rate=1.07%, first attempt success rate=85.9%). Among these devices, OFOI and NFOI most likely required multiple intubation attempts, while the other devices had relatively high rates of success on the first intubation attempt.\n\nWhen comparing our results with those obtained by Ezri et al.2, the most striking difference is the increased use of video laryngoscopes. Ezri et al., reported fiberoptic intubation and the LMA as the most popular in management of the difficult airway; no data was reported on the utilization rates of video laryngoscopes. The results of a similar survey completed by Canadian Anesthesiologists were recently presented at the Society of Airway Management Meeting 2013, where Mehta et al.6 showed that the preferred alternative airway technique in difficult intubation situations was video laryngoscope. In a 2005 survey5 the same authors found that the preferred devices were lighted stylet, bronchoscope, and intubating laryngeal mask airway (Table 1).\n\nThere has been a rapid acceptance of video laryngoscopy as an important technique in the management of difficult airway situations. It is our opinion though, that while video laryngoscopy is preferred for ease of use and a faster learning curve, the technique of flexible fiberoptic endoscopy offers invaluable advantages: nasal and oral intubation, double lumen tube or bronchial blocker placement for thoracic surgery, therapeutic bronchoscopy, and it is preferred for awake technique intubation. The device versatility also makes it economical not to mention the greater value of education and training of future anesthesiologists.", "appendix": "Author contributions\n\n\n\nDC, PVK, and CAH initiated the study. RC and PVK performed the analysis. RC and RAC wrote the abstract and main body of the article. DC supervised the process. All authors critically edited the correspondence and agreed to the final content.\n\n\nCompeting interests\n\n\n\nDavide Cattano has received grant support from Karl Storz Endoskope Inc. Dr Cattano serves on the speaker bureau for Cadence and is a paid consultant for Smiths Medical. Dr Carin A. Hagberg has received grant support from AMBU, Covidien, and Karl Storz Endoskope Inc. She serves on the speaker bureau for Covidien, LMA North America, and Ambu A/S.\n\n\nGrant information\n\nThe study was sponsored by an educational grant from the Foundation for Anesthesia Education and Research (FAER). Davide Cattano has received grant support from Karl Storz Endoskope Inc. Dr Carin A. Hagberg has received grant support from AMBU, Covidien, and Karl Storz Endoskope Inc.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary materials\n\nPre-operative assessment form and post-operative evaluation used in the survey on alternative airway devices at the Memorial Hermann Hospital – Texas Medical Center at Houston, TX, USA.\n\nPre-operative Airway Assessment Form pdf file.\n\nPost-operative Evaluation pdf file.\n\n\nReferences\n\nMerli G, Guarino A, Petrini F, et al.: Should we really consider to lay down the Macintosh laryngoscope? Minerva Anestesiol. 2012; 78(9): 1078–9. PubMed Abstract\n\nEzri T, Szmuk P, Warters RD, et al.: Difficult airway management practice patterns among anesthesiologists practicing in the United States: have we made any progress? J Clin Anesth. 2003; 15(6): 418–22. PubMed Abstract \| Publisher Full Text\n\nJenkins K, Wong DT, Correa R: Management choices for the difficult airway by anesthesiologists in Canada. Can J Anesth. 2002; 49(8): 850–6. PubMed Abstract \| Publisher Full Text\n\nBokhari A, Benham SW, Popat MT: Management of unanticipated difficult intubation: a survey of current practice in the Oxford region. Eur J Anesthesiol. 2004; 21(2): 123–7. PubMed Abstract\n\nWong DT, Lai K, Chung FF, et al.: Cannot intubate-cannot ventilate and difficult intubation strategies: results of a Canadian national survey. Anesth Analg. 2005; 100(5): 1439–46. PubMed Abstract \| Publisher Full Text\n\nMehta A, Tam A, Yau B, et al.: Preferences of Canadian Anesthesiologists in difficult intubation and cannot intubate – cannot ventilate situations. Abstracts of Oral Presentations – Society of Airway Management Meeting. 2013. Publisher Full Text" }	[ { "id": "4807", "date": "30 May 2014", "name": "Sairam Parthasarathy", "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\nExcellent and insightful information about approaches to intubation", "responses": [] }, { "id": "5327", "date": "11 Jul 2014", "name": "Suzanne Karan", "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 represents a growing body of research which will hopefully inform the appropriate education and training of our residents. The authors present data regarding the reported use of a variety of devices to manage the difficult airway. The venue of using F1000Research to quickly and more widely disseminate this information is highly valuable. The granularity of knowing user rate and first attempt success rate for these devices is a constructive addition that should be incorporated in future surveys to allow for comparison. With more data, it will be interesting to note whether the lower first attempt success for FOI (nasal or oral) stays the same or even drops compared with video laryngoscopy as the latter becomes more prevalently used and taught. The authors are commended in adding their research findings and their thoughtful opinions for review.", "responses": [] }, { "id": "5326", "date": "16 Jul 2014", "name": "Ronald Pearl", "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 report from one of the leading institutions in airway management confirms the rapid growth in the use of alternative airway devices, especially video laryngoscopes, but emphasizes that, on one hand, the majority of patients are still intubated using direct laryngoscopy, and, on the other hand, there remains an important role for fiberoptic intubation. The report documents a large number of available alternative airway devices but does not address the issues of how many different devices are required for the potential range of airway issues and how many devices can the standard practitioner be trained to use and maintain competency in their use.", "responses": [] } ]	1	https://f1000research.com/articles/3-114
https://f1000research.com/articles/3-112/v1	16 May 14	{ "type": "Observation Article", "title": "Visual record of intertidal disturbance caused by drift ice in the spring on the Atlantic coast of Nova Scotia", "authors": [ "Willy Petzold", "Maike T. Willers", "Ricardo A. Scrosati", "Willy Petzold", "Maike T. Willers" ], "abstract": "In the early spring of 2014, an unusually large amount of sea ice drifted from the Gulf of St. Lawrence, where it had been produced, towards the open Atlantic Ocean through the Cabot Strait, between Nova Scotia and Newfoundland, Canada. In early April, significant amounts of drift ice reached the Atlantic coast of mainland Nova Scotia. The ice floes persisted in those coastal waters for up to 16 days, depending on the location. During that time, the ice fragments caused extensive physical disturbance in rocky intertidal communities, removing high quantities of seaweeds and invertebrates. For example, at a location where the ice stayed for 9 days, the loss of macroalgal and invertebrate biomass was almost total. At a location where the ice stayed for 4 days, losses were lower, albeit still high overall. Such a magnitude of disturbance is not common on this coast, as sea ice had not reached the surveyed locations in the previous 4–5 years. We suggest that the frequency of ice scour events may help to predict intertidal community structure. This notion could be tested through multiannual surveys of ice conditions and biological communities along the Atlantic coast of Nova Scotia.", "keywords": [ "The NW Atlantic coast exhibits cold-temperate conditions. As with similar systems in other parts of the world", "the distribution and abundance of rocky intertidal species are greatly influenced by latitudinal changes in temperature and pelagic food supply1–3. Unlike most other temperate coastal systems", "however", "on the NW Atlantic coast", "sea ice may affect considerably the survival of intertidal species and", "consequently", "the structure of biological communities." ], "content": "Observation\n\nThe NW Atlantic coast exhibits cold-temperate conditions. As with similar systems in other parts of the world, the distribution and abundance of rocky intertidal species are greatly influenced by latitudinal changes in temperature and pelagic food supply1–3. Unlike most other temperate coastal systems, however, on the NW Atlantic coast, sea ice may affect considerably the survival of intertidal species and, consequently, the structure of biological communities.\n\nWhile a stable ice coverage of intertidal habitats (the ice foot) prevents benthic organisms from experiencing very low temperatures during low tides4, the movement of ice fragments because of tides, currents, winds, and waves can severely damage or remove intertidal organisms5,6. On many NW Atlantic shores from relatively enclosed bodies of water, such as gulfs or bays, sea ice readily develops on the sea surface every winter, causing a great deal of disturbance in rocky intertidal communities when ice fragments move around7. On the open Atlantic coast, however, ice does not form on the sea surface. Nonetheless, drift ice produced in enclosed bodies of water may still reach the open coast and cause damage there. Such is the case of the open Atlantic coast of Nova Scotia. Between mid-winter and early spring, sea ice produced in the large Gulf of St. Lawrence often drifts towards the Atlantic Ocean through the Cabot Strait, between Nova Scotia and Newfoundland (Figure 1). The floating ice fragments then move southwards along the Atlantic coast. The extent to which the ice floes travel south varies between years, often being limited but reaching the central coast of mainland Nova Scotia in unusually extreme years8 (Canadian Ice Service).\n\nThe coastal locations from mainland Nova Scotia referred to in the text are indicated with black dots. The arrows indicate the direction that the sea ice originated in the Gulf of St. Lawrence normally follows when drifting out of the gulf. The asterisk shows the southernmost reach of the drift ice on the coast of mainland Nova Scotia in 2014, according to the Canadian Ice Service.\n\nIn 2014, a large amount of floating ice fragments came out of the Gulf of St. Lawrence between late winter and early spring. In its travel south along the Atlantic coast, the ice came in contact with an approximately 92-km-long stretch of coastline in mainland Nova Scotia (Figure 1). Ice fragments varied widely in size, but together formed a relatively compact coverage of the sea surface (Figure 2–Figure 3). Such a high influx of sea ice eventually devastated rocky intertidal communities. Before the arrival of the ice in early April, intertidal habitats were abundantly covered with seaweeds and invertebrates. For example, in Whitehead (45° 12' 43.5\" N, 61° 10' 25.6\" W, Figure 1), high and middle intertidal elevations from wave-exposed habitats exhibited a well developed canopy of Fucus algae (Figure 4) and an abundance of mussels (Mytilus) and barnacles (Semibalanus balanoides) in understory habitats (Figure 5). At middle and low elevations from wave-exposed habitats in Tor Bay Provincial Park (45° 10' 57.6\" N, 61° 21' 19.4\" W, Figure 1), a dense canopy of Chondrus crispus (a red alga) dominated the landscape, while, at the lowest intertidal elevations, kelp (mostly Laminaria and Saccharina) formed a conspicuous canopy that covered smaller algae, such as C. crispus and coralline algae, and a diversity of small invertebrates (Figure 6).\n\nPicture taken at low tide in the afternoon of 3 April 2014 at a wave-exposed site in Whitehead, showing a full coverage of the intertidal zone by seaweed canopies and the drift ice approaching the shore. The sea surface was calm on that day.\n\nPicture taken at low tide in the late afternoon of 3 April 2014 from the wave-exposed site in Whitehead shown in Figure 2. This picture shows the variable size of the ice fragments at the time of their first contact with the shore.\n\nPicture taken at low tide on 3 April 2014 at the wave-exposed site from Whitehead shown in Figure 2. This picture shows the intertidal zone covered by a Fucus canopy at high and middle elevations (f) and by Chondrus crispus and coralline algae at low elevations (c), which also exhibit the first ice fragments that contacted the shore on that day.\n\nPicture taken at low tide on 3 April 2014 at the wave-exposed site from Whitehead shown in Figure 2. This picture shows the mussels and barnacles that were abundant in understory habitats below the Fucus canopy, which was removed to take the picture.\n\nPicture taken at low tide on 4 April 2014 at a wave-exposed site in Tor Bay Provincial Park, showing a well developed canopy of Chondrus crispus at middle-to-low elevations (c) and a kelp canopy at the lowest elevations (k). The little plates that are visible above the C. crispus zone were drilled into the rocky substrate to study barnacle recruitment. The sea surface was calm on that day, and sea ice was visible towards the horizon.\n\nThe ice scour that occurred on those shores for days until the ice melted removed a large amount of algae and invertebrates. The duration of the presence of sea ice on the shore was related to the intensity of biological damage. For instance, in Whitehead, which sustained 9 full days (between 4–12 April) of ice coverage (Canadian Ice Service), the intertidal zone underwent an almost total loss of organisms (Figure 7). At Tor Bay Provincial Park, which sustained 4 days (between 6–9 April) of ice coverage (likely because it is farther away from the ice source), biomass losses were also high (Figure 8), but some organisms were able to survive in some protected areas (Figure 9). The magnitude of ice scour in mainland Nova Scotia in 2014 was such that ice effects were even observed in wave-sheltered habitats. In such habitats, which are normally dominated by the perennial brown seaweed Ascophyllum nodosum9, the movement of ice fragments is relatively limited5. However, in 2014, biomass losses were still high in some wave-sheltered habitats, leaving extensive areas without any significant macroalgal coverage (Figure 10).\n\nPicture taken at low tide on 30 April 2014 at the wave-exposed site from Whitehead shown in Figure 4, showing the extreme removal of algae and invertebrates by the sea ice, which stayed for 9 days on the shore. The little barnacle recruitment plates visible in this picture were drilled to the rocky substrate at an elevation of approximately 2/3 of the full intertidal range (between chart datum, or 0 m in elevation, and the elevation where the barnacles located highest on the shore occurred before the ice scour).\n\nPicture taken at low tide on 27 April 2014 at the wave-exposed site from Tor Bay Provincial Park shown in Figure 6. This picture shows the almost complete loss of the macroalgal cover shown in Figure 6 because of the effects of ice scour.\n\nPicture taken at low tide on 27 April 2014 at the wave-exposed site from Tor Bay Provincial Park shown in Figure 6. This picture shows the post-ice survival of some algae in protected sites.\n\nPicture taken at low tide on 27 April 2014 at a wave-sheltered site in Tor Bay Provincial Park, showing the loss of the Ascophyllum nodosum canopy that had previously covered these habitats for an undetermined number of years (at least 10, based on observations by R.A.S.). Remains of A. nodosum canopies are seen in the upper-left corner and upper-right corner of this picture.\n\n\nConcluding remarks\n\nAs the duration of the ice presence on the open Atlantic coast of Nova Scotia generally decreases from the Cabot Strait southwards, albeit not linearly (Canadian Ice Service), the observations herein described suggest that intertidal community structure may be influenced by latitude mediated by ice scour effects. We predict that communities from northern locations in this coastal range would remain in early successional stages, as such places receive drift ice from the Gulf of St. Lawrence mostly every year. Conversely, communities from southern locations in this coastal range might reach more mature stages because of sea ice failing to reach those places for a number of years. This notion is supported by the fact that, on Sober Island (44° 49' 20.3\" N, 62° 27' 26.5\" W), which is located south of the southernmost reach of the sea ice in 2014 (Figure 1) and has not been exposed to ice floes since 2007 (Canadian Ice Service), intertidal communities were well developed and seaweeds extensively covered the rocky surface shortly after the 2014 ice season (Figure 11). We suggest that a multiannual survey of ice conditions and biological communities along the open Atlantic coast of Nova Scotia could reveal the ecological role that sea ice plays on intertidal community organization in this cold-temperate coastal system.\n\nPicture taken at low tide on 1 May 2014 at a wave-exposed site in Sober Island, showing a full coverage of seaweed canopies, as the sea ice had not reached this shore during the previous 7 years.", "appendix": "Author contributions\n\n\n\nWP, MTW, and RAS all participated in the field surveys. RAS wrote the manuscript and WP and MTW provided critical comments to produce the final version.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe field surveys were funded by a Discovery Grant (# 311624) awarded to RAS by the Natural Sciences and Engineering Research Council of Canada (NSERC).\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\nAdey WH, Hayek LAC: Elucidating marine biogeography with macrophytes: quantitative analysis of the North Atlantic supports the thermogeographic model and demonstrates a distinct subarctic region in the northwestern Atlantic. Northeast Nat. 2011; 18(mo8): 1–128. Publisher Full Text\n\nCole SWB, Scrosati RA, Tam JC, Sussmann AV: Regional decoupling between NW Atlantic barnacle recruit and adult density is related to changes in pelagic food supply and benthic disturbance. J Sea Res. 2011; 65(1): 33–37. Publisher Full Text\n\nTam JC, Scrosati RA: Mussel and dogwhelk distribution along the north-west Atlantic coast: testing predictions derived from the abundant-centre model. J Biogeogr. 2011; 38(8): 1536–1545. Publisher Full Text\n\nScrosati R, Eckersley LK: Thermal insulation of the intertidal zone by the ice foot. J Sea Res. 2007; 58(4): 331–334. Publisher Full Text\n\nScrosati R, Heaven C: Field technique to quantify intensity of scouring by sea ice in rocky intertidal habitats. Mar Ecol Prog Ser. 2006; 320: 293–295. Publisher Full Text\n\nJohnson LE: Ice scour. In: Denny MW, Gaines SD editors; 2007; Encyclopedia of Tidepools & Rocky Shores, University of California Press, Berkeley, USA. Reference Source\n\nScrosati R, Heaven C: Spatial trends in community richness, diversity, and evenness across rocky intertidal environmental stress gradients in eastern Canada. Mar Ecol Prog Ser. 2007; 342: 1–14. Publisher Full Text\n\nMinchinton TE, Scheibling RE, Hunt HL: Recovery of an intertidal assemblage following a rare occurrence of scouring by sea ice in Nova Scotia, Canada. Bot Mar. 1997; 40: 139–148. Publisher Full Text\n\nWatt CA, Scrosati RA: Regional consistency of intertidal elevation as a mediator of seaweed canopy effects on benthic species richness, diversity, and composition. Mar Ecol Prog Ser. 2013; 491: 91–99. Publisher Full Text" }	[ { "id": "4812", "date": "23 May 2014", "name": "Mathieu Cusson", "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 observation article provides a portrait of an ice scouring event on the Nova Scotia Atlantic shoreline. The article is very short yet explains and demonstrates well the devastating effects of ice scouring. The authors provide several pictures that efficiently illustrate the disturbed community. It would have been ideal if the authors had taken pictures 4 and 7 as well as 6 and 7 with a common guide mark (for scale) on the ground. However, the pictures are clear enough to appreciate ice scouring impacts on the benthic communities. Such impacts on macrobenthic communities from ice scouring are very common in the St. Lawrence estuary and gulf. I do agree that communities in the latter regions would remain in early successional stages on exposed substrates.All information provided in this communication is accurate. I consider the observations reported by this paper as interesting and useful.I would suggest the authors add the reference below into the paragraph 2, first sentence:Bergeron P, Bourget E (1986) Shore topography and spatial partitioning of crevice refuges by sessile epibenthos in an ice disturbed environment. Mar Ecol Prog Ser 28:129-145", "responses": [] }, { "id": "4930", "date": "29 May 2014", "name": "Gregorio Bigatti", "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 work registered the disturbance caused by drift ice on benthic intertidal communities in rocky shores of Nova Scotia. The authors documented in real time the ice coverage and the latter disturbance of the intertidal algae and invertebrates. It is recommendable to measure the biomass losses if this phenomenon occurs again in the zone. The observations made here are useful to other researchers working in similar habitats, opening new questions on successional stages and the ecological role of sea ice plays on intertidal community organization of benthic communities in Nova Scotia.I suggest including a map of northern America in Figure 1, and an inset with the location of Nova Scotia and ice drift (actual fig 1).", "responses": [] } ]	1	https://f1000research.com/articles/3-112
https://f1000research.com/articles/3-111/v1	14 May 14	{ "type": "Case Report", "title": "Acute respiratory distress syndrome secondary to Mycobacterium abscessus lung infection – a case report", "authors": [ "Santhosh Gheevarghese John", "Tirdad T. Zangeneh", "Sairam Parthasarathy", "Tirdad T. Zangeneh", "Sairam Parthasarathy" ], "abstract": "Mycobacterium abscessus (M. abscessus) is among the most common rapidly growing mycobacteria causing infections in humans. Skin and soft tissue infections, especially those following trauma or surgery are the most common infections caused by this pathogen. We describe the case of a 22-year old female with cerebral palsy and chronic respiratory insufficiency requiring nocturnal ventilation through a tracheostomy, who was admitted to a hospital with worsening shortness of breath, fever, and ventilator dependence. The patient was diagnosed with acute respiratory distress syndrome (ARDS) caused by bilateral pneumonia due to M. abscessus infection. The patient received prolonged antibiotic treatment and respiratory support which led to clinical recovery and bacteriological cure.", "keywords": [ "A 22 year old Hispanic woman who had a history of cerebral palsy", "seizure disorder with tracheostomy and required night-time mechanical ventilation for chronic respiratory insufficiency", "was admitted to the hospital with a gradually worsening shortness of breath", "fever", "chills and rigor that persisted over the course of 1 week." ], "content": "Case report\n\nA 22 year old Hispanic woman who had a history of cerebral palsy, seizure disorder with tracheostomy and required night-time mechanical ventilation for chronic respiratory insufficiency, was admitted to the hospital with a gradually worsening shortness of breath, fever, chills and rigor that persisted over the course of 1 week.\n\nThe patient had been on assist control ventilation through a tracheostomy and portable home-ventilator (Pulmonetics LTV 1000; Carefusion Corp. San Diego, CA) for 12 months prior to this hospitalization. Although she had suffered from frequent respiratory tract infections in the past, with the last one being 4 months prior to this admission, she required mechanical ventilation only at night-time, as she breathed spontaneously during the day, and received feeding through a gastro-enterostomy tube. There was no previous history of seizures. The patient’s family denied any recent weight loss, night sweats or skin rashes. Her home medications included levetiracetam 250 mg per G tube twice daily, multivitamins and inhaled albuterol as needed for bronchospasm. Her vital signs upon admission were as follows: blood pressure of 84/50 mmHg, heart rate of 109 beats per minute, respiratory rate of 20 breaths per minute, oral temperature 38.3°C, and oxygen saturation of 98%. Her home ventilatory setting included assist-control mode at a rate of 10/minute, tidal volume of 400 ml and positive end-expiratory pressure of 5 cm H2O through a positive end-expiratory pressure (PEEP) valve built into the lambda ventilator circuit. She was immediately admitted to the intensive care unit (ICU).\n\nPhysical examination revealed a frail, non-verbal female with habitus and posture suggestive of spasticity, connected to a ventilator via tracheostomy. Physical examination of the chest revealed scattered crepitations over bibasilar lung fields, and predominantly over the right axillary area. She had severe spasticity affecting both the upper and lower extremities. Her initial chest X-ray showed a multifocal pneumonia predominantly confined to the right upper lobe. Admission labs revealed hemoglobin 12.5 gm/dL, white count 8.8 × 103/L, platelets 197 × 103/L. Her serum sodium levels were 138 mEq/dL, and the blood urea nitrogen (BUN) and creatinine values were 20 mg/dL and 0.9 mg/dL respectively. Arterial blood gas showed a pH of 7.30 and PaO2 of 70 mmHg.\n\nIn addition to fluid resuscitation with normal saline infusions, given her history of pneumonia caused by Pseudomonas aeruginosa 3 months before the hospitalization, the patient was started on intravenous piperacillin and tazobactam at 3.75 g every 8 hours. Her blood cultures were drawn and were subsequently reported as negative for both bacterial and fungal organisms. Respiratory cultures obtained from tracheal aspirate grew P. aeruginosa sensitive to meropenem and hence she was started on meropenem 500 mg intravenously every 6 hours, after discontinuing piperacillin and tazobactam as well as vancomycin. Despite being on antibiotics for 7 days and showing negative culture for Pseudomonas, she continued to be febrile with temperature spikes of 38.6°C with night sweats. Oxygen saturation progressively worsened and subsequent chest X-rays showed a progressive increase in pulmonary infiltrates bilaterally, involving all four lung quadrants. Further laboratory cultures for respiratory pathogens including a viral panel, Legionella, Mycoplasma and cocci serology were negative. Her sputum obtained from the tracheal aspirate grew Mycobacterium abscessus, initially identified as M. chelonae-abscessus complex, on day 10 of admission to the ICU. By this time, the patient showed a progressive decline in her respiratory status with computerized tomographic (CT) scans of the chest revealing findings consistent with a ‘tree-in-bud’ appearance on the initial images (Figure 1). Intravenous cefoxitin 2 g every 6 hours along with intravenous azithromycin 500 mg daily was started. Her progressive deterioration in respiratory status and worsening of the pulmonary infiltrates were consistent with acute respiratory distress syndrome (ARDS) characterized by bilateral four quadrant infiltrates and PaO2/FiO2 ratio <200 in the absence of features of left atrial hypertension (Figure 2). A 2D echocardiogram showed left ventricular ejection fraction of 65% without any evidence of atrial or ventricular dysfunction. The patient required a fractional inspired oxygen (FiO2) concentration of 100%, PEEP of 10 cm H2O, and low tidal volume ventilation with permissive hypercapnia. She then received low tidal volume “lung protective” ventilation keeping the tidal volume of 400 ml and careful fluid management to prevent fluid overload for the underlying ARDS. The patient was also maintained on respiratory support therapy including chest physical therapy and inhaled bronchodilators including albuterol.\n\nThere is also a trace left pleural effusion.\n\nThere are also bilateral small pleural effusions.\n\nOn day 15 of hospital admission, the drug sensitivity report revealed that the infection was caused by a multidrug resistant M. abscessus strain, sensitive only to tigecycline and amikacin. Therefore the patient was started on intravenous (IV) tigecycline 50 mg twice daily and inhaled amikacin 500 mg twice daily. After being on these regimen for 7 days, there was a progressive improvement in oxygenation and radiological evidence for resolution of pulmonary infiltrates. After a prolonged hospital stay for 6 weeks, the patient was subsequently discharged to a long-term acute care facility and inhaled amikacin for 4 weeks and IV tigecycline for a total of 12 weeks. Sputum cultures were repeated at the 8th and 12th week of treatment with tigecycline and revealed no mycobacterial growth (“bacteriological cure”). The patient’s bacteriological cure from this atypical mycobacterium correlated with her clinico-radiological course, as evidenced by the improvement in the pulmonary infiltrates. A subsequent chest CT scan that was performed 8 weeks after commencement of tigecycline confirmed these findings (Figure 3). She returned to her home ventilator settings of assist-control mode with minimal oxygen requirements (fractional inspire oxygen of 30%). Subsequent outpatient follow-up at 3 months up revealed a stable cardiorespiratory status at her baseline.\n\nThere are bilateral residual trace pleural effusions.\n\n\nDiscussion\n\nM. chelonae-abscessus complex, subspecies abscessus accounts for 85% of the pulmonary diseases caused by M. chelonae isolates1. M. abscessus is a rapidly growing, aerobic acid fast bacillus (AFB) that produces non-pigmented colonies on most types of solid medium in less than 7 days2. Although M. abscessus is a rapid grower, complete speciation of the organism is important but not often done in most clinical laboratory settings. Hence, M. abscessus is often confused with M. chelonae. The organism isolated from our patient was initially identified as M. chelonae but was subsequently confirmed as M. abscessus. This is important for several reasons as both organisms have different clinical manifestations. M. chelonae typically affects patients who are immunosuppressed and are on chronic steroid therapy. The majority of these patients present with disseminated skin and soft tissue infections, predominantly affecting the extremities3. However M. abscessus is more nosocomial and virulent and it causes disease in patients regardless of their immune status. Regarding antibiotic sensitivity, both M. chelonae and M. abscessus are resistant to most antibiotics, except to clarithromycin and amikacin3. M abscessus is usually sensitive to cefoxitin whereas cefoxitin sensitivity of M. chelonae is variable3. Our patient was empirically started on cefoxitin based on this clinical observation. However, M. abscessus isolated from our patient was resistant to most antibiotics except tigecycline and amikacin. In fact the patient showed a rapid improvement of her symptoms and clinico-radiological findings after being started on tigecycline and amikacin. Subsequent sputum cultures were negative for M. abscessus.\n\nPatients with localized M. abscessus infections respond well to appropriate treatment. The treatment approach to disseminated disease varies and is typically based on in vitro susceptibility testing. The goals of therapy should be more realistic and should include symptomatic improvements, radiographic regression of infiltrates, improvements in sputum culture positivity and conversion to negativity. A combination therapy of amikacin and cefoxitin or imipenem for 2 to 4 weeks is recommended, although the cost of therapy and morbidity has an effect on the treatment outcome. In cases of failure to treat with the above regimen or in cases of drug resistance, drugs showing efficacy in vitro against M. abscessus, such as linezolid, tigecycline (a tetracycline derivative) and telithromycin (a ketolide) have been shown to be effective4. M. abscessus isolated from our patient was resistant to most of the drugs and hence was treated with tigecycline. The patient presented a significant clinical, radiological improvement and also had a negative sputum conversion. Although M. abscessus species are rapid growers, the usual clinical course is indolent. To the best of our knowledge, no cases of rapid progression of pulmonary disease to ARDS secondary to M. abscessus have been reported in the literature. Fulminant, rapidly progressive diseases with M. abscessus have been associated with gastroesophageal disorders and cystic fibrosis5. The diagnosis of M. abscessus is based on the diagnostic criteria formulated by the American Thoracic Society4. Although not considered as confirmatory tests, imaging studies and especially high resolution CT scans (HRCT) play an important role in the diagnosis and follow-up of M. abscessus infection. Non-tuberculous mycobacterial (NTM) infections are classified in two groups based on HRCT presenting either a cavitary pattern or a nodular bronchiectasis pattern with tree-in-bud appearance6. M. abscessus cases show widely scattered tree-in-bud appearance on CT chest7. Our patient presented the same characteristic appearance on her chest CT scan. Small centrilobular nodules of soft tissue attenuation interlinked to linear structures of similar size originating from a single stalk gives the characteristic tree-in-bud appearance on imaging studies8. Although this pattern was classically reported in mycobacterial infections, it has also been described in other infectious, inflammatory, immunologic and pulmonary vascular disorders9. Differential diagnosis of tree-in-bud appearance on CT scans in the setting of acute rapidly progressive disease should also include bacterial pneumonia, Haemophilus influenza infection, tumor emboli in the pulmonary vasculature and inhalation of toxic fumes9. Upon identifying this pattern on HRCT scans; work-up for the broad differentials should include a thorough medical history and clinical examination. The histopathological features contributing to this pattern in mycobacterial infection are the accumulation of caseous material within or around the bronchioles, with stalk and terminal tufts, a manifestation of caseous material in the terminal bronchioles and alveolar ducts, respectively8. The tree-in-bud sign is important in discriminating between phases of unfavorable progression and phases of quiescence or resolution10. Chest CT scan is important when surgical intervention is planned or when assessing the effect of chemotherapy, as the resolution of the infection can be demonstrated sooner and quickly by radiological evidence.\n\nIn summary, our case report highlights the importance of identifying M. abscessus as a cause of severe respiratory failure requiring intense respiratory support and aggressive medical management. We also emphasize the importance of monitoring drug sensitivity in these cases as it would improve the chances of successful treatment of these potentially fatal infections.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient’s mother.", "appendix": "Author contributions\n\n\n\nAll authors have seen and approved the text of the manuscript and taken responsibility for its contents. Dr John has collected the relevant clinical data and drafted the article which was critically revised and edited by the co-authors, Dr Zangeneh and Dr Parthasarathy.\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\nSingh N, Yu VL: Successful treatment of pulmonary infection due to Mycobacterium chelonae: case report and review. Clin Infect Dis. 1992; 14(1): 156–161. PubMed Abstract \| Publisher Full Text\n\nJeon K, Kwon OJ, Lee NY, et al.: Antibiotic treatment of Mycobacterium abscessus lung disease: a retrospective analysis of 65 patients. Am J Respir Crit Care Med. 2009; 180(9): 896–902. PubMed Abstract \| Publisher Full Text\n\nMueller PS, Edson RS: Disseminated Mycobacterium abscessus infection manifesting as fever of unknown origin and intra-abdominal lymphadenitis: case report and literature review. Diagn Microbiol Infect Dis. 2001; 39(1): 33–37. PubMed Abstract \| Publisher Full Text\n\nGriffith DE, Aksamit T, Brown-Elliott BA, et al.: ATS Mycobacterial Diseases Subcommittee; American Thoracic Society; Infectious Disease Society of America. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med. 2007; 175(4): 367–416. PubMed Abstract \| Publisher Full Text\n\nGriffith DE, Girard WM, Wallace RJ Jr: Clinical features of pulmonary disease caused by rapidly growing mycobacteria: an analysis of 154 patients. Am Rev Respir Dis. 1993; 147(5): 1271–8. PubMed Abstract \| Publisher Full Text\n\nPolverosi R, Guarise A, Balestro E, et al.: High-resolution CT of nontuberculous mycobacteria pulmonary infection in immunocompetent, non-HIV-positive patients. Radiol Med. 2010; 115(2): 191–204. PubMed Abstract \| Publisher Full Text\n\nKurashima A: [Radiographic findings of pulmonary nontuberculous mycobacteriosis other than Mycobacterium avium complex]. Kekkaku. 2009; 84(8): 577–83. PubMed Abstract\n\nRossi SE, Franquet T, Volpacchio M, et al.: Tree-in-bud pattern at thin-section CT of the lungs: radiologic-pathologic overview. Radiographics. 2005; 25(3): 789–801. PubMed Abstract \| Publisher Full Text\n\nBastawrous S, Hirschmann JV: A 71–year-old man with fever, productive cough, and tree-in-bud pattern on chest CT scan. Chest. 2013; 144(2): 700–703. PubMed Abstract \| Publisher Full Text\n\nFerrara I, Cappabianca S, Brunese L, et al.: HRCT in detection of pulmonary infections from nontuberculous mycobacteria: personal experience. Radiol Med. 2009; 114(3): 376–389. PubMed Abstract \| Publisher Full Text" }	[ { "id": "4783", "date": "09 Jun 2014", "name": "David E Griffith", "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\nComment #1: The authors correctly note that fulminant pneumonia due to M. abscessus to date has been limited to a very few clinical circumstances such as lipoid pneumonia and cystic fibrosis. To attribute fulminant pneumonia to M. abscessus in an entirely new setting requires rather rigorous \"proof\". For several reasons, I am not sure that \"proof\" exists for this case.a) The patient's course was consistent with bacterial pneumonia followed by adult respiratory distress syndrome. b) The patient had pseudomonas isolated from respiratory secretions and had multiple previous pneumonias presumably due to pseudomonas.\n\nc) Although tigecycline has less activity against pseudomonas than some other gram negative rods, it still has some activity and certainly amikacin has significant anti-pseudomonal activity. d) The radiographic findings were non-specific (as noted by the authors). e) The source of the M. abscessus in respiratory specimens could have been the tracheostomy site. f) The patient had rapid radiographic and microbiologic response, both unusual for M. abscessus lung disease with any antibiotic combination. g) The total length of treatment was very short for M. abscessus lung disease, and mostly with one agent. h) A biopsy would have been helpful for establishing the diagnosis. i) Is it possible the patient had ongoing aspiration (a factor that might account for some of the radiographic findings but might also support the mycobacterial diagnosis)?Comment #2: I think the discussion is too ambitious. In my opinion the focus of the discussion should be persuading the reader about the veracity of the claim that M. abscessus caused this fulminate clinical course.a) The discussion about separating M. chelonae from M. abscessus in the lab is important, but can be summarized by the urgent need for widespread adoption of molecular methods not currently available in most mycobacteriology labs. There is no excuse M. abscessus to be reported as part of a complex. To make things even more interesting, there appear to be 3 subspecies of M. abscessus (abscessus, massiliense and boletti). This problem is clearly peripheral to the thrust of the paper. b) The authors do not mention the major functional (and clinical) difference between M. chelonae and M. abscessus which is that M. abscessus has a functioning inducible macrolide resistance gene (erm gene) whereas M. chelonae does not. This means that M. chelonae isolates are usually macrolide susceptible while M. abscessus organisms are not. Again, this observation is somewhat peripheral to the main point of the manuscript. c) Overall, I think the discussion could be shortened with more focus on the main thrust of the manuscriptComment #3: Some minor comments:a) The authors discuss treatment duration for M. abscessus infections as though the organism dictates the duration of therapy, rather duration of therapy is strongly dictated by the site of the infection. Duration of therapy differs between skin and soft tissue infection, lung disease, bone and joint disease, etc. b) In the case report, I would like to see the pCO2 and HCO3 levels and the anion gap.", "responses": [] }, { "id": "5049", "date": "27 Jun 2014", "name": "Hartmut Lode", "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 case report on an ARDS probably induced by M. abscessus is an interesting clinical observation and broadens the etiologic spectrum of infections involved in ARDS.The case is in general well described and the discussion addresses the current knowledge and problems with this pathogen. However, the first reviewer Dr. Griffith has already performed a critical review of this paper and has raised strong concerns for acceptance of the diagnosis of an infection by M. abscessus. Dr. Griffith is an outstanding expert in the field of nontuberculous mycobacterial infections and I do agree with most of the concerns raised by Dr. Griffith. However, the critical point of the case is whether Pseudomonas was the etiologic pathogen or not:The authors are reporting that after one week of treatment with 3 times 500 mg meropenem daily, the patients’ clinical findings worsen. - It would be helpful to know the body weight of the patient (sufficient meropenem doses?) and also some biomarkers of inflammation (start of therapy versus day 7), to have some additional objective data for the non-response. It would also be useful to include any data on the control of the trachea flora, concerning bacteria, on day 7.Dr. Griffith has reservations in accepting the relatively fast response to the M. abscessus directed therapy, but even in more localized M. abcessus infections the median time until sputum conversation was only 1 month (Jeon K et al.;AJRCCM 2009).In summary, the authors should give more information to demonstrate that Pseudomonas was not the etiologic pathogen as outlined above. The authors should also make a revision of the paper following all the proposals of Dr. Griffith.", "responses": [] }, { "id": "5514", "date": "31 Jul 2014", "name": "Michael Loebinger", "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 case of respiratory infection leading to ARDS which they suggest is secondary to Mycobacterium abscessus infection. They make this assumption based on the growth of M. abscessus from a tracheal aspirate in addition to the lack of improvement on meropenem, improvement on tigecycline and amikacin and the radiological findings. This is an interesting case however the main issue is that the evidence that M. abscessus is the microbe responsible for the clinical course is unconvincing. A bacterial pneumonia followed by ARDS would be more likely in view of the previous history and timescale, and the radiology is non-specific as detailed by the previous reviewers. In addition the treatment regimen was short and non-standard for M. abscessus pulmonary infection and the response more rapid than would be expected.\n\nIt appears that M. abscessus was obtained from a single sputum (via tracheal aspirate) sample only which would be insufficient to meet diagnostic criteria for NTM infection. Repeated growths of this organism or a bronchoscopic sample would have been helpful in this case. A history of frequent respiratory tract infections is mentioned. It would be important to know if any samples had been sent for mycobacterial culture previously and in addition the number of subsequent samples post treatment.\n\nIn its present state we do not feel that the authors have convincingly demonstrated ARDS secondary to M. abscessus lung infection as per the title and would suggest that more evidence would be required before this case report is suitable for indexing.", "responses": [] } ]	1	https://f1000research.com/articles/3-111

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.